Moe: Competitive Landscape and Competitor Ratings

A genuinely good 14-year-old self-funded B2B inventory SaaS, and the clearest evidence in this set that a consumer catalog-your-stuff app only survives by leaving consumer.

What it is

Sortly is inventory and asset tracking software for small and mid-sized businesses: photograph an item, barcode it, set alerts, report, integrate. It sells into construction, medical, warehouse, electrical, interior design and education (sortly.com/pricing, 2026-09-16). Harmonic classifies it B2B, 94 employees.

It began on the consumer side. Its Crunchbase slug is still "my-things-app" and the iOS app dates to 2012-06-05, when it was a home-inventory product. That pivot is why it belongs in this set: the longest-running experiment on the consumer "photograph everything you own" idea, and its answer was to sell to contractors.

Founders and team

Dhanush Balachandran, Founder and CEO since 2013-09 (Harmonic person 81279). BE in Electronics and Communications, NIT Tiruchirappalli. MBA, Indian School of Business. Lead Engineer at Tejas Networks 2000 to 2003, Strategic Marketing Manager at Intel 2004-10 to 2012-01, iOS Lead at Jiff 2012-09 to 2013-09. Technical by training and shipped work.

Worth flagging: Harmonic records him as Software Engineering Manager at DJI from 2015-08 to 2017-05, overlapping his Sortly tenure by 21 months, so Sortly was a side project through its third and fourth years. He has now been on it 13 years. Harmonic lists no co-founder. A Chief Revenue Officer was hired 2024-07, and headcount runs 32 engineering, 12 product, 10 data, 9 marketing, 7 sales, 6 design.

Checks run: Harmonic company and person records, site and pricing page, App Store listing and seller entity, web search on founder and company. No adverse record found in the checks run.

Product and traction

US App Store, 2026-09-16: 9,481 ratings at 4.73 average, released 2012-06-05, last updated 2026-09-16, the same day this card was written. Harmonic web traffic 506,700 monthly, LinkedIn 16,758 followers.

9,481 ratings over 14 years is slower than Dupe.com's 5,828 in ten months, but this is a paid B2B tool so ratings understate it badly. The 94-person headcount and 506,700 monthly visits say this is a real operating company with meaningful revenue.

Business model and unit economics

Self-serve subscription: free tier plus Advanced $49/mo, Ultra $149/mo, Premium $299/mo and Enterprise, all at 50 percent off under a running "Flash Sale" banner (sortly.com/pricing, 2026-09-16). Standard SMB SaaS, with the low churn of being the system of record for a contractor's tool inventory.

Harmonic records $306,000 across two events, the last debt in May 2020. A 94-person company that has consumed $306,000 of outside capital is funding itself out of revenue, or Harmonic's record is materially incomplete. I report the Harmonic figure per the data rule and flag both readings.

Position in this market

Sortly is the cautionary tale and the survivor at once. Every consumer scan-and-catalog product in this set chases the primitive it started with in 2012, and it reached the conclusion the graveyard implies: consumers will not pay to catalog their possessions, businesses will.

Base decomposition

• Market and competition: 6. SMB inventory and asset tracking is durable, fragmented and pays real money. Asset Panda and EZOfficeInventory compete but nobody owns it.
• Product and wedge: 6. Fourteen years of depth, barcode scanning, labelling, alerts, reporting, integrations, a release shipped today.
• Traction: 6.5. 9,481 ratings at 4.73, 94 employees, 506,700 monthly visits, published prices to $299 a month. A revenue business.
• Unit economics: 6. Self-serve SMB SaaS with an enterprise tier, supporting 94 people on $306,000 of outside capital.
• Team: 5.5. Technically trained founder, 13 years on the problem. No co-founder on record, and a 21-month overlap with a full-time DJI job early on.
• Founder trust and disclosure: 5. Fully public pricing, which most of this set does not offer. The permanent "Flash Sale" banner is a mild dark pattern. Nothing adverse found.

Mean = (6 + 6 + 6.5 + 6 + 5.5 + 5) / 6 = 35 / 6 = 5.83. No adjustment. Base 5.83.

Formula

5.83 Base + 1.00 EV ($306,000, the $0-$1M tier) + 1.00 Geo (USA) + 0.00 Chain = COMPANY 7.83.

The EV bonus does a lot of work here, and it should: 94 employees on $306,000 is a real property of the business. But a 7.83 on a 14-year-old company is not a 7.83 on an investable seed opportunity.

For Daxos

PASS, and it is not close. Harmonic shows currently_raising null and the last financing event was debt in May 2020. A profitable 94-person SaaS in its fourteenth year is not selling $100k to $500k of seed equity, and if it sells anything it will be to a private equity buyer at a revenue multiple. There is no entry price because there is no entry, and Daxos adds nothing: no capital needed, no introductions needed, customers outside the fund's network.

Sortly's value to Daxos is as a reference point. It is the answer to "what happens if a consumer scan-and-catalog product works": selling $149-a-month software to electrical contractors, 94 people, 14 years in, having never raised a real round. A good business and a bad venture return. Fit 1.5 out of 10.

Threat to Moe: 1. Different customer, different product, no overlap.

What could not be checked

Revenue, ARR, customer count, churn and profitability are undisclosed. Whether the $306,000 figure is complete could not be verified; the record may simply be thin, and Harmonic names neither investor. Google Play installs were not visible. Any early co-founder, and whether a seed round went uncaptured, could not be established.

back to the ranked table

The best asset in this batch by a wide margin, a bootstrapped Canadian TCG portfolio app with 64,990 iOS ratings and 2.9M monthly web visits, which is exactly why it does not need Daxos money.

What it is

Collectr is a portfolio tracker for trading card games. Scan a card, it identifies the exact set, number and variant and adds it to your collection; the app then values the whole collection daily from real sales data, charts it over time, shows cost basis and gains, runs a price guide with five years of raw and graded history, analyses whether a proposed trade is fair, and wraps a social layer of showcases and creators around it (getcollectr.com and getcollectr.com/pro, both fetched 2026-09-16). Twenty-plus games, from Pokemon and Magic to One Piece and Lorcana.

It does not list anything anywhere. There is no marketplace and no cross-posting. It sits in the vertical-with-liquid-comps lane, next to CollX and Alt, not in the seller-tool lane.

Founders and team

Three co-founders, all from 2021-11, and they have worked together before, which is the single most reassuring fact on this card.

Adam Hijleh, CEO. Queen's University, software design, 2008 to 2013. Software engineer then senior engineer at Flipp 2016 to 2018, technical lead at Coinsquare, Director of Innovation and Cloud Architecture at Finastra 2020 to 2021, then engineering leadership on membership growth at Uber from 2021-11 to 2023-04. Co-founder and CTO of Doorr 2018 to 2020 (Harmonic person 27574701). Technical founder, unambiguously.

Abbas A., CPO. Toronto Metropolitan University, computer science. Senior product owner at Flipp 2013 to 2017, Director of Product Management at Finastra 2020 to 2023, co-founder and CPO of Doorr 2018 to 2020 (Harmonic person 437340).

Muhammad Rashid, COO. Toronto Metropolitan University, BCom in IT and management. Seven years at Flipp through 2017, Director of BD at Filogix, GM Mortgage at Properly, co-founder and CEO of Doorr 2018 to 2020 (Harmonic person 161788602).

Doorr went through Techstars in the Spring 2019 cohort, per the record of Petre Neytchev, who was principal engineer there and joined Collectr as Principal Engineer in 2025-10 after Index Exchange, Shakepay and eleven years at IBM, with a Waterloo BMath in computer science. So the core group has shipped one venture-backed company together already and pulled their old principal engineer back in.

One thing to raise directly with them: all three co-founders show a concurrent current role at Selby Lane Capital beginning 2023-04 (CTO, Head of Product, Head of Integrations), and the COO also shows a current mortgage agent role at Vogue Mortgages from the same date. That may be stale LinkedIn data carried into Harmonic, but on its face it reads as part-time attention split across two ventures. It is the first question I would ask.

Checks run: Harmonic company record and four Harmonic person records, company site and PRO pricing page, App Store lookup and public App Store page for id 1603892248, Harmonic social links. No adverse record found in the checks run.

Product and traction

iOS app "Collectr - TCG Collector App," id 1603892248, seller Collectr Inc, released 2022-01-31, version 2.5.7, last updated 2026-09-12, 64,990 ratings at a 4.88 average as of 2026-09-16 (itunes.apple.com/lookup, 2026-09-16). That 4.88 is the highest average in this whole competitive set. Android is live.

Harmonic web traffic 2,925,000. Instagram 368,646 followers, X 26,143, Facebook 31,000, LinkedIn 898. The site claims over 100,000 reviews across platforms and "millions" of users.

For scale inside this batch: Collectr has more iOS ratings on its own than every funded seller tool in the set combined, and it did it with no recorded funding. The nearest independent comparables on the same date are PSA at 56,837, CollX at 48,085, PriceCharting at 26,463 and HoloDex at 22,673. Collectr leads them all.

Business model and unit economics

Collectr PRO is $7.99 monthly or $59.99 annual (App Store in-app purchase list, 2026-09-16; the site advertises the annual as $4.99 a month equivalent). PRO buys unlimited scanning, five-plus years of pricing history, advanced filters, export, custom backgrounds, a PRO badge and partner perks. There is also merchandise.

The structure is sound: a large free base doing the scanning, a soft paywall on scan volume and data depth, and a creator and community layer that raises retention. The missing piece is transactions. With portfolio positions, cost basis and a trade analyser, Collectr already knows what every user owns and wants, which is the raw material for a marketplace take rate it is not yet charging. That is the upside case and also the risk, since building liquidity means taking on TCGplayer, eBay and Whatnot.

Scan inference costs money against a free tier this large, so the PRO conversion rate is the number that decides whether 11 people is a comfortable headcount or a tight one.

Position in this market

This is the shape the category brief says actually compounds: a vertical with liquid comps, a real catalogue, and an audience that returns weekly because their collection has a price that moves. It is upstream of a marketplace rather than downstream of a platform API, so it carries none of the cross-lister platform-ban risk that defines the tool lane. Its exposure is different and milder: TCGplayer and eBay could bundle portfolio tracking, and Next Vision's FoilSnap is already shipping a cheap version of the scanner, at 13,375 ratings in twelve months.

Base decomposition

Market and competition 6.0: large, growing, liquid, with credible independent share already taken. Product and wedge 6.0: scan plus portfolio plus trade analyser plus social is a coherent system, not a feature. Traction 7.0: 64,990 ratings at 4.88, 2.9M monthly web visits, 368k Instagram, all with no recorded funding. Unit economics 5.0: real subscription at $59.99 a year, but conversion unknown, scan costs real, and no transaction revenue yet. Team 6.5: three repeat co-founders with a prior Techstars company, an ex-Uber engineering leader as CEO, and a senior principal engineer they had worked with before. Founder trust and disclosure 4.0: concurrent Selby Lane Capital roles for all three founders since 2023-04, unexplained.

Mean of 6.0, 6.0, 7.0, 5.0, 6.5, 4.0 is 5.75. No adjustment. Base 5.75.

Formula

5.75 base + 1.00 EV + 0.75 geo + 0.00 chain = 7.50 company.

The EV bonus is the mechanical result of a $0 Harmonic record. Read it carefully: 11 employees and 2.9M monthly visits are being paid for somehow, so either subscription revenue covers it or there is money Harmonic has not captured. Verify before relying on the bonus.

For Daxos, price and verdict

Harmonic shows currently_raising null and no round in five years. This is the best business in the batch and the one least likely to want a $100k to $500k cheque. If it ever raises, the entry would be priced off 2.9M monthly visits and a paying subscriber base, which puts it well above anything in this category that Daxos has seen, and above the $2.0M post that the scan-and-list wedge has cleared. Daxos adds little to a Toronto team that has already done Techstars.

Worth a conversation anyway, for two reasons: it is the clearest read available on what real consumer demand in this category looks like, and the concurrent-roles question could reveal a founder group that is ready to go full time with capital. Daxos fit 5.5. MEET.

What could not be checked

Revenue, PRO conversion rate, paying subscriber count, Google Play scale, churn, whether Selby Lane Capital roles are current or stale, whether any unrecorded angel or revenue-based financing exists, and any marketplace plans.

back to the ranked table

A 19-year bootstrapped sold-price database across games, cards, comics, coins and Funko with 26,463 iOS ratings and four employees, the real moat in this market, and almost certainly unbuyable at a seed price.

What it is

PriceCharting is the reference price guide for collectible categories with liquid secondary markets: video games across every Nintendo, Sega, Atari, PlayStation and Xbox platform in NTSC, PAL and Japanese variants, plus Pokemon, Magic, Lorcana, One Piece and YuGiOh cards, sports cards, comics, coins, Funko Pops and LEGO. The about page states 45,000-plus video games priced (pricecharting.com/page/about, read 2026-09-16).

Around that data it has built a collection tracker, a lot value calculator, a photo appraiser, an eBay Lot Bot, an eBay Game Sniper, a public API, an iOS and Android app, and a peer-to-peer marketplace with buy, sell and newest-listings surfaces.

Founders and team

JJ Hendricks and Michael Hendricks, brothers. JJ compiled pricing data for his own video game retail business, then the two made it public in 2007 (pricecharting.com/page/about). JJ's LinkedIn lists him as founder at PriceCharting.com and SportsCardPro.com, and he runs a public YouTube channel under @jjgames. Harmonic reports headcount 4.

Checks run: Harmonic enrich on pricecharting.com (no person records returned), the site root and /faq, /offers, /sell and /page/about, the App Store lookup for app id 6452190948, and a web search on the founder name. The public blog at blog.pricecharting.com runs continuously from at least 2009 and doubles as a changelog.

No adverse record found in the checks run. No litigation, dissolved entity, bankruptcy or disputed claim surfaced. Disclosure is the strongest in this batch: named founders, a real about page, terms of service, a return policy, published API documentation, a 17-year public blog, and an explicit statement of how the company earns money.

Both founders are operators rather than career software people, but the product is a data pipeline and a marketplace they have run themselves for 19 years, so I record TECHNICAL_FOUNDER as yes on shipped evidence.

Product and traction

26,463 US iOS ratings at 4.82, from a 2023-09-11 release, last updated 2026-09-10 and checked 2026-09-16. That is the highest rating count in this batch, and it was earned by an app that is only three years old on top of a 16-year-old website, so it understates the underlying audience.

Peer context from the same App Store queries, all checked 2026-09-16: Collectr 64,990 ratings, PSA Card Scanner 56,837, CollX 48,085, Sports Card Investor 37,718, HoloDex 22,673, TCGplayer 14,579. PriceCharting sits mid-pack among the card scanners while being the only one that also covers games, comics, coins, Funko and LEGO.

Business model and unit economics

Three lines, all disclosed. Affiliate commissions on eBay and Amazon click-throughs, stated in the site footer. A PriceCharting Pro subscription, price not published on the pages I could read. API access. There is also a paid ad-removal option ("Don't Want Ads?").

The marketplace charges nothing. JJ Hendricks has stated no listing fees, no selling fees and no buying fees, with no plans to charge sales fees, monetizing instead through eBay and Amazon affiliate links (medium.com/@hendricks_jj media kit and reviewfix.com interview, 2016, via search 2026-09-16).

Four people, 19 years, no outside capital, still shipping. The efficiency is exceptional. The deliberate zero take rate is the whole story: this company built the demand-side data that would let it charge, and chose not to.

Position in this market

The category brief says the tool lane has a hard ceiling and that where the thesis compounded it was vertical with liquid comps. PriceCharting is the purest example of the second half of that sentence, and it got there without venture money at all. It owns the sold-price history that the newer scanner apps in this research effectively resell. Several of them cite "millions of real sales" without saying whose.

It is also the clearest illustration of why this market does not produce venture returns. The most defensible asset in the entire competitive set is a four-person bootstrapped business in Denver that refuses to charge a take rate.

Base decomposition

• Market and competition 5.0. Vertical collectibles with liquid comps, the one corner of this market that works, though the card niche specifically is crowded by better-funded apps with larger rating counts.
• Product and wedge 6.0. Nineteen years of sold-price history across six collectible categories is the only genuine moat found anywhere in this research, and it cannot be cloned in a weekend.
• Traction 6.5. 26,463 iOS ratings at 4.82, 45,000-plus games priced, continuous operation since 2007, still shipping in September 2026.
• Unit economics 4.5. Bootstrapped and durable on affiliate, subscription and API with four people, but a deliberate 0 percent marketplace fee forgoes the revenue line that would make it scale, and affiliate income depends on eBay's program terms.
• Team 5.5. Two brothers, 19 years on one problem, started from an actual game retail business, deep domain credibility.
• Founder trust and disclosure 6.5. Named founders, public blog and changelog since 2009, published about page, terms, return policy and API docs, and an explicit honest statement of how it makes money.

Mean = (5.0 + 6.0 + 6.5 + 4.5 + 5.5 + 6.5) / 6 = 34.0 / 6 = 5.67. Adjusted -0.17 because a business that has spent 19 years refusing to charge a marketplace fee has structurally chosen against the growth path. Base = 5.50.

Formula

5.50 Base + 1.00 EV ($0 raised) + 1.00 Geo (USA, Denver) + 0.00 Chain = COMPANY 7.50.

The 7.50 measures the quality of the business, not its availability. Two of the three points come mechanically from having raised nothing and being American.

For Daxos

No check, but take the meeting. PriceCharting has been profitable and bootstrapped for 19 years, is owned entirely by two brothers, has never raised and has no reason to start. If they ever wanted liquidity it would be a strategic or private equity deal priced off nearly two decades of cash flow, not a $100k to $500k seed check at a seed price. Daxos cannot add distribution, data or hiring help to a company that already has all three. Any entry price a seed fund could reach here would not exist.

The conversation is still worth having, and for a specific reason. JJ Hendricks sits on the sold-price data that most of the scanner apps in this research quote, which means he can see the true demand under every one of them. That is the single best available read on whether anything in this category is worth backing. DAXOS 4.0, VERDICT MEET, as a diligence call rather than a deal.

What could not be checked

Revenue, profit, PriceCharting Pro subscription price, API pricing, marketplace GMV, site traffic and app MAU are all undisclosed. Harmonic returns no person records, no valuation and a founding date of 2021-03 that contradicts the company's own 2007 history, so I used the primary source instead. The /page/fees and /page/api-documentation URLs returned a shell page or a 404 through the fetch tools, so the zero-fee terms come from the founder's own published media kit and a 2016 interview rather than from a current fee schedule, and the fee policy may have changed since.

back to the ranked table

Fourteen-year-old B2B field documentation platform for insurance restoration contractors, 119 people and 3,000-plus shops, adjacent to consumer resale only through its contents-inventory module, and far past any seed entry.

What it is Encircle sells field documentation software to property restoration contractors, the crews that dry out and rebuild after water and fire losses. The app captures photos, video, 360 images, notes, moisture readings and equipment logs on site, then generates insurance-carrier-ready reports and, since the 2026 "Scope to Estimate" release, an IICRC-aligned scope that exports into Xactimate (getencircle.com, 2026-09-16). The site claims 3,000-plus restoration shops (getencircle.com, 2026-09-16). The only piece that touches this competitor set is the Contents module: photograph items during a packout, get AI-generated descriptions and per-item records, produce an inventory report. That is item capture with a valuation attached, but the buyer is a contractor and the payer is an insurer. There is no resale, no listing, no marketplace.

Founders and team Paul Donald, CEO and co-founder since 2012. Harmonic's person record shows BlackBerry VP ISV Alliances 1999-2003 and VP CDMA 2003-2009, then an entry reading "Google, COO, 2010-05 to 2011-02" (Harmonic person 406425, pulled 2026-09-16). A nine-month COO tenure at Google is not plausible as written and is almost certainly a mis-parsed regional or divisional title carried over from LinkedIn; I flag it rather than repeat it as fact. The BlackBerry decade in Waterloo is consistent and is the real credential.

Christophe Biocca, co-founder 2012-08 to 2023-03, University of Waterloo Bachelor of Mathematics in Computer Science (Harmonic person 406429). Technical, and gone for three years.

Ronuk Raval, "Chief Person" from 2012-08, University of Waterloo Bachelor of Computer Science, two Facebook software engineering internships in 2011 and 2012 (Harmonic person 406426). Effectively the technical lead of the founding group.

Christopher Austin carries a co-founder title from 2014-05 to 2020-01 as "Curriculum Developer" with a sociology MA and a career of English instruction in Japan (Harmonic person 13774517). This is a late, non-core use of the founder label, worth noting but not damning.

Checks run: Harmonic company and person records, the company site, App Store publisher data, general web search on the company name. No adverse record found in the checks run. LinkedIn profile scraping was not run.

Product and traction The iOS app has 129 ratings at 3.77 average, released 2013-02-28, last updated 2026-09-16 (iTunes lookup 604527488, 2026-09-16). That count is low, but for a crew tool installed by a manager it is a weak proxy, unlike the consumer apps in this set. Headcount of 119 and a 14-year history are the better signals. LinkedIn followers 6,097 (Harmonic, 2026-09-16).

Business model and unit economics Priced on annual job volume, not seats. Field Documentation is $285 per month for up to 200 jobs a year; Scope to Estimate is $583 per month, currently discounted 15 percent; 12-month contracts, unlimited users and storage (getencircle.com/pricing, 2026-09-16). Unlimited seats on a volume meter is the right shape for this buyer, and moving from documentation into estimating is a real ACV expansion, since Xactimate estimating is where the money in the claim actually sits.

Position in this market Encircle is in this competitor set only by adjacency. It sits in the insurance restoration workflow, a lane with a paying commercial buyer, a compliance-driven reason to document, and no dependence on marketplace APIs. That makes it structurally healthier than the cross-lister lane, where the platforms give the loop away free and break the integrations. But it also means it tells Daxos nothing about the consumer resale question and poses no threat to a consumer scan-and-list product.

Base decomposition Market and competition 6.0: a real, defensible B2B niche, though Verisk and Xactimate own the estimating standard Encircle must plug into. Product and wedge 6.5: documentation to floor plan to scope to Xactimate is a genuine workflow lock, not a feature. Traction 6.5: 3,000-plus shops claimed and 119 employees are real, but 14 years to get there is slow. Unit economics and business model 6.0: volume-metered annual contracts with unlimited seats, healthy shape, unverified margins. Team 7.0: two Waterloo CS founders plus a BlackBerry telecom executive, though one technical co-founder left in 2023. Founder trust and disclosure 5.0: the Harmonic capital record is materially incomplete and one founder's headline prior role does not survive a plain reading. Mean of 6.0, 6.5, 6.5, 6.0, 7.0, 5.0 is 37.0 divided by 6 equals 6.17. Adjusted down 0.25 for the 14-year clock and a capitalization no external record can reconstruct. Base 5.90.

The formula 5.90 + 0.75 (EV, Harmonic-recorded money of roughly $1.565M including the 2022 grant it cites) + 0.75 (Canada) + 0.00 (chain) = 7.40. That score is soft. The EV bonus is being paid on a funding record Harmonic itself contradicts. A company with 119 employees and 3,000 customers did not get there on $65,000, and if the true total exceeds $10.9M this is out of scope entirely.

For Daxos, price and verdict DAXOS 1.5. There is no seed entry here. Encircle is 14 years old, has 119 employees, sells to commercial contractors, and last touched the public capital record with a government grant in 2022. It is not raising a round a $100k to $500k check could join, the price would be a growth-stage secondary at best, and Daxos adds nothing to a restoration software company in Kitchener. PASS.

What could not be checked Actual capitalization, revenue, net retention, and whether the 3,000-shop figure is paying customers or total accounts. Harmonic's funding record is internally inconsistent and no substitute figure was taken from the web, per the data rules.

back to the ranked table

The strongest team in the batch and the only one building a demand side rather than just a tool, but the marketplace is embryonic and the seed is already closed with a priced A coming above Daxos range.

What it is

An AI selling app. Snap a photo, Hero prices the item off eBay sold comps and Facebook Marketplace data, writes the listing, and cross-posts to eBay and Facebook Marketplace (herostuff.com and App Store description, 2026-09-16). Video and audio input are supported, so a seller can describe an item by voice instead of typing. Two things sit alongside the app and matter more than the app: a free SEO price guide covering collectibles, cards, electronics and fashion, and a consumer-facing Shop at herostuff.com/shop that lists items for sale across 25 categories. The CTO's own one-line description is "Vision AI meets ecommerce marketplace" (Harmonic person 160661490).

Founders and team

This is the best-credentialed team in the competitor set, by a wide margin.

Joshua Voydik, CEO and co-founder since 2024-01, University of Oregon, business. Co-founded Swell, an API-first headless ecommerce platform, 2019-05 to 2022-04, which his own record says raised $25M from Bonfire Ventures and VMG (Harmonic person 350418). Before that, Head of Partnerships at Headphones.com and partnerships at CartStack. Commercial rather than engineering, but with ecommerce infrastructure operating history.

Jeremy Herrman, co-founder and CTO since 2023-10, University of Pittsburgh BS. Twenty years shipping software: BxVideoSolutions 2004 to 2008, software engineer at Carnegie Mellon's Software Engineering Institute 2008 to 2011 as the first engineer on the Immersive Learning Technologies team, UPMC telemedicine architecture 2011 to 2013, founder of Plethora 2013 to 2019, co-founder of Volition 2019 to 2022, then Applied AI at Outerbase 2023, which Cloudflare acquired (Harmonic person 160661490). Deeply technical with venture-backed founder history.

Kai Gradert, co-founder and designer since 2024-04. Senior Designer at Google 2005 to 2008. Design lead for Adobe Lightroom Mobile, hired with the single objective of shipping it on iPhone and iPad. Design Lead at AngelList 2014 to 2020, joining as the third designer, covering AL Venture, AL Talent and later Product Hunt. Co-founder and design lead at Cheers 2010 to 2013, which pivoted to Flipagram and was acquired by ByteDance. Founding designer at August Health 2020 to 2024 (Harmonic person 19077). One prior exit, and an unusually strong consumer design record for this category.

Checks run: Harmonic company record and three person records, company site including shop, price guide, business and pricing pages, App Store search and lookup, one general web search on the founders and their prior companies covering shutdowns and litigation. No litigation, dissolved entity, disputed claim or inflated title surfaced. No adverse record found in the checks run. One open item: Plethora's outcome is not confirmed, and a podcast episode titled "Leaving a Startup and Lessons Learned from Plethora" with co-founder Nick Pinkston (manufacturinghappyhour.com, episode 114) suggests it did not end well. Herrman left in 2019-08. I did not verify this.

Product and traction

US App Store id 6479718293, checked 2026-09-16: 2,060 ratings, 4.72 average, released 2024-09-20, version 2.28.0 shipped 2026-09-12. Roughly half of SellRaze's rating count over a similar window, on a later start. Site claims 85,000-plus sellers and over 1 million scans and listings, and says the app is rated on Google Play as well as Apple, though no Play numbers are given (herostuff.com, 2026-09-16). Headcount 25 (Harmonic). Coverage is narrow: eBay and Facebook Marketplace only, against SellRaze's five-plus and Vendoo's full set.

Business model and unit economics

Three tiers, monthly, Stripe checkout. Starter $14.99 with unlimited pricing scans but only 10 cross-posts a month and 100 studio credits. Seller $29.99 with 150 cross-posts and 500 credits. Business $89.99 with unlimited cross-posts and 1,500 credits (herostuff.com/pricing, 2026-09-16). A business page pitches bulk catalog import from Shopify, WooCommerce or a spreadsheet.

The metering choice is deliberate and risky. Scans are unlimited on every paid plan while cross-posts are capped, which means the vision inference cost, the expensive part, is uncapped on a $14.99 floor while the cheap part is rationed. That is a pricing structure built for adoption, not margin. It only makes sense if scans are a funnel into something that monetizes later, which is where Shop comes in.

Position in this market

Hero is the only company in this batch making the move the category brief says is the one that works. The venture outcomes here are demand networks, not seller tools, and Hero is assembling one: a free price guide to own the "what is this worth" search intent, an app that converts that intent into supply, and a Shop to capture the demand side and eventually a take rate. That is a coherent path past the tool-lane ceiling and no one else in this batch has one.

It is also very early. The Shop inventory on 2026-09-16 was mostly low-value collectibles, action figures and toys priced $8 to $35, which is the hardest possible cold start. Building liquidity against Facebook Marketplace, Mercari, Depop and Vinted from a standing start is the thing almost nobody does. Meanwhile the tool half carries the same platform dependence as everything else in the lane: eBay's 2026-02 user agreement bans LLM-driven agentic bots, and there has never been a public Facebook Marketplace C2C listing API, so Hero's two integrations are exactly the two most exposed.

Base decomposition

Market and competition 3.5: the tool-lane ceiling still applies, with a small premium for aiming past it at the demand side. Product and wedge 5.5: the funnel architecture of free price guide into app into shop is the best-designed in the batch; marketplace coverage is the narrowest. Traction 5.0: 2,060 ratings at 4.72 in two years and 85,000 claimed sellers is solid but half the leader. Unit economics 4.0: unlimited scans on a $14.99 floor leaves inference cost uncapped; no take rate yet. Team 7.0: Adobe and AngelList design lead with a ByteDance exit, a CEO who co-founded a $25M-raised ecommerce platform, a CTO with twenty years shipping and a Cloudflare-acquired employer. Founder trust and disclosure 5.0: numbers consistent across site, store and Harmonic; nothing adverse found; one unverified prior-company outcome.

Mean = (3.5 + 5.5 + 5.0 + 4.0 + 7.0 + 5.0) / 6 = 30.0 / 6 = 5.00. Adjust +0.25 because the demand-side strategy is the only credible route past this category's ceiling present anywhere in the batch. Base = 5.25.

The formula line

COMPANY = 5.25 Base + 1.00 EV + 1.00 Geo + 0.00 Chain = 7.25

Flag on the arithmetic. Harmonic records funding_total as $0.00 because the 2025-11 seed amount is undisclosed, which mechanically lands Hero in the $0 to $1M tier and hands it the maximum +1.00 EV bonus. Twenty-five headcount in San Francisco with FJ Labs, Long Journey and Mento on the cap table does not happen on under $1M. If the round was $3M to $5M, the correct EV bonus is +0.50 and COMPANY is 6.75. Treat 6.75 as the honest number and 7.25 as the formula's output on the data available.

For Daxos, price and verdict

Hero closed its seed in 2025-11 and is not raising now. At 25 people it will be raising a Series A inside twelve months, and a team with this pedigree building a marketplace will price that round far above what a $100k to $500k cheque can enter. There is no obvious allocation here at any point in the next year.

The reason to engage anyway is informational and optional. Hero is the one company in this batch testing whether a listing tool can bootstrap into a demand network, and that experiment is the single most valuable thing happening in this market for anyone deciding whether to be in it at all. If Shop starts showing real liquidity, the A becomes a genuine question rather than a price problem. If it does not, Hero is a very well-built company in a lane with a mid-eight-figure ceiling, and the answer is no.

VERDICT: MEET. Take the call, ask specifically about Shop gross merchandise value and repeat buyers, and treat the answer as a read on the whole category. Do not expect to get into this round.

What could not be checked

The 2025-11 seed amount, valuation and terms: undisclosed, Harmonic records $0. Revenue, paid conversion and churn: no source. Shop gross merchandise value, listing count, buyer count or whether any take rate is charged: no source. Google Play ratings and installs: claimed on the site, not published. Whether eBay and Facebook integrations use official APIs or browser automation. Plethora's outcome and Herrman's role in it.

back to the ranked table

YC W25 bid-ask marketplace for graded Pokemon cards with real volume on $500k, but the sole technical founder left in February 2026.

What it is

Misprint runs a real-time bid and ask marketplace for graded and sealed Pokemon trading cards. Buyers and sellers post two-sided quotes instead of fixed list prices, and machine learning fills in a reference price from comparable sales. The homepage shows a running counter of $83,628,669 in cumulative market volume and advertises platform fees "as low as 2.5%" plus a paid subscription that it says saves members $376 a year (misprint.com, checked 2026-09-16). There is a card scanner, an iOS app and an Android app.

Founders and team

Eva Herget is CEO. University of Chicago economics, Promontory Investment Research, then equity analyst at Goldman Sachs July 2023 to July 2024 (Harmonic person 170396366, 2026-09-16). She left Goldman after building a Pokemon card flipping business; the YC page says she reached $500k ARR in three months doing it and was selling over $40k a month at $16k to $20k monthly profit before founding the company (ycombinator.com/companies/misprint, 2026-09-16). Genuine domain operating history, not a thesis.

Jonathan Jenkins was technical co-founder and CTO. Penn State PhD in mathematics, machine learning doctoral research, Carnegie Mellon teaching assistant before that. His own record describes building the entire data and ML platform as sole technical founder. Harmonic shows that role ending 2026-02, with Jenkins now a founding ML engineer at Crustdata (Harmonic person 193889700, 2026-09-16). For a company whose whole claim is ML pricing, losing the only person who built the pricing engine is the most important fact on this card.

The replacement is William Thomas: first technical hire March 2025, Head of Engineering January 2026, CTO September 2026. University of Pittsburgh dual bachelor's in mathematics and computer science, then roughly three years as a software engineer at DICK'S Sporting Goods (Harmonic person 21780848, 2026-09-16). Competent, shipped payments and checkout, but a second-job engineer replacing a machine learning PhD is a downgrade in technical depth.

Checks run: Harmonic company and person records, YC company page, Crunchbase and PitchBook profile links via search, company site. No adverse record found in the checks run: no litigation, no dissolved entities, no disputed claims, and the YC traction figures line up with the Harmonic funding record.

Product and traction

iOS app 987 ratings at 4.84 average, released 2025-03-17, last updated 2026-09-14 (iTunes lookup 6741099582, 2026-09-16). Harmonic web traffic 189,000. Headcount 4. In this market the comparable free scanner apps sit far higher: Collectr 64,990 ratings, CollX 48,085, PriceCharting 26,463, Card Ladder 9,602 (iTunes, 2026-09-16). Misprint is a transactional marketplace rather than a free utility, so ratings undercount it, but 987 is still early. The $83.6M cumulative volume figure is self-reported on the site and cannot be independently verified.

Business model and unit economics

Take rate as low as 2.5% plus a subscription. At 2.5% the $83.6M lifetime volume implies roughly $2.1M of cumulative gross take. Payment processing on collectibles runs close to 3%, so unless the buyer carries the fee, the advertised rate leaves almost nothing. Positioning on "lowest fees" against TCGplayer and eBay at roughly 10% to 13% wins share and forfeits margin at the same time. The subscription is the more interesting line and is the one to probe.

Position in this market

Misprint is on the right side of the structural divide. It is a demand network in a vertical with liquid comps, which is the only shape in this market that has compounded. It is not a seller tool sitting on top of other people's platforms, so the platform dependence risk that kills cross-listers does not apply. What does apply is competitive density: TCGplayer inside eBay, Whatnot, Fanatics Collect, Collectors and PSA all want the same volume, and all of them are far better capitalised.

Base decomposition

Market and competition 4.5: large and liquid, but the most crowded corner of collectibles and every large competitor is funded at another order of magnitude. Product and wedge 5.5: a true two-sided order book is a real difference from list-price marketplaces, and the scanner plus apps are shipped. Traction 5.5: 987 ratings and $83.6M claimed volume in eighteen months on $500k is strong capital efficiency, though the volume number is self-reported. Unit economics 4.0: a 2.5% take against roughly 3% payment cost is not obviously a business yet. Team 4.0: excellent domain operator as CEO, but the technical founder is gone and the successor is junior. Founder trust and disclosure 5.5: public metrics, YC record and Harmonic agree, no adverse findings.

Mean = (4.5 + 5.5 + 5.5 + 4.0 + 4.0 + 5.5) / 6 = 29.0 / 6 = 4.83. Adjust +0.25 for sitting in the demand-network lane rather than the tool lane. Base = 5.10.

Formula

5.10 + 1.00 + 1.00 + 0.00 = 7.10.

For Daxos, price and verdict

Harmonic records no round since the $500k pre-seed of February 2025 and does not flag the company as raising. The pre-seed implies an estimated $2.0M post (Harmonic ESTIMATED, from typical 25% dilution). Nineteen months and $83.6M of claimed volume later, a YC seed with these numbers would likely price somewhere between $12M and $25M post. A $250k check at that price buys one to two percent of a business whose realistic ceiling is a trade sale to eBay, Fanatics or Collectors in the tens to low hundreds of millions. That does not return a fund, and Daxos adds nothing a YC network does not already supply. Verdict MEET, not IN: the CEO is a real operator with real volume and the conversation is worth an hour, but two questions gate any check, namely who actually owns the pricing model now that Jenkins has gone, and what the blended net take rate is after payments.

What could not be checked

The $83.6M market volume, GMV, revenue and active buyer counts are all self-reported and unverified. No Google Play install figure was visible from the site. The current cap table, any post-pre-seed SAFE and whether a priced seed is in market are all unknown. Jenkins's departure terms and whether he retains equity or any ongoing advisory role could not be established.

back to the ranked table

Thirty-year bootstrapped Dutch collection-database company, 15 people, 24,425 iOS ratings at a 4.85 average, entirely subscription funded and entirely uninvestable at seed.

What it is CLZ, formerly Collectorz.com, makes collection database software for comics, movies, books, music and games, plus barcode and cover scanner companions. The core motion is recognition: point the camera at a comic cover and the app identifies the item against CLZ Core, its own metadata database, filling in series, issue number, variant, creators, characters, cover art and key-issue flags (App Store, 2026-09-16). Users add grade, storage box, purchase price and custom fields.

Harmonic: founding_date 2000, headcount 15, funding_total $0 across zero rounds, VENTURE_UNKNOWN, Consumer B2C, Amsterdam (Harmonic enrich, 2026-09-16), no person records. The site sits behind Cloudflare and returned 403 to every fetch, so detail below comes from App Store listings and company history.

Founders and team Alwin Hoogerdijk, sole founder and president. He studied Electrical Engineering and Computer Science in Delft and was a programmer at SIG in Utrecht from 1994. In 1996 he wrote the first Windows product, Music Collector, to catalog his own CDs, added DVD, book, comic and game products in 2000 and 2001, then moved to mobile and web (clz.com blog and alwinhoogerdijk.com, 2026-09-16). He wrote the original code and has been on this problem for thirty years.

No CTO or lead engineer is named apart from the founder, and the LinkedIn page carries 132 followers (Harmonic socials, 2026-09-16), which for a fifteen-person company signals a team that does not recruit rather than trouble.

Checks run: Harmonic company record (zero person records), App Store publisher and app lookups, web search on the founder and company history, direct fetch of clz.com and clz.com/about (both 403). No adverse record found. LinkedIn scraping was not run.

Product and traction US App Store, pulled 2026-09-16. Comics 14,764 at 4.86 (rel 2009-06-29, upd 2026-08-24), Movies 3,845 at 4.84 (rel 2009-05-10), Books 2,612 at 4.82, Games 1,569 at 4.81, Music 1,365 at 4.84, all updated 2026-08-20; CLZ Barry barcode scanner 193 at 4.72, last upd 2022-11-14; CLZ Scanner 77 at 4.68, last upd 2025-06-18. Total 24,425.

The 4.8-plus average across the five main apps is the highest quality signal in this comparison set, and four shipped updates within the last month. Seventeen years of continuous App Store presence with current releases is an execution record nothing else here can show.

Business model and unit economics Straight subscription. CLZ Comics is $1.99 per month or $19.99 per year with a seven-day trial, plus an optional CovrPrice comic-values add-on at $90 per year (App Store, 2026-09-16). That add-on is the most interesting number here: CLZ resells third-party pricing data at four and a half times the price of its own app and finds takers. A narrow willingness-to-pay datapoint for "what is this worth" in a vertical with liquid comps. Fifteen salaried people funded by $19.99-a-year subscriptions for thirty years means the model works, and that the ceiling is visible from the ground.

Position in this market CLZ is the oldest survivor in the adjacent lane and the cleanest illustration of its economics. It owns the asset everyone else in the scan-and-value category is building, a thirty-year metadata database with working cover recognition, and has converted it into fifteen jobs in Amsterdam and nothing more. It is also outflanked where it matters: Collectr, free, launched 2022-01-31, already carries 64,990 ratings against CLZ's 24,425 across seven apps, and owns trading cards, the one vertical with real resale volume. CLZ never built one. There is no listing, cross-posting, take rate or marketplace relationship in the product, so it threatens no resale product.

Base decomposition Market and competition 4.0: a durable niche with a defensible data asset, but free vertical apps take the categories that carry money. Product and wedge 6.5: cover recognition into a proprietary database across five verticals, 4.85 average, updates shipped this month. Traction 6.5: 24,425 ratings and fifteen employees paid out of revenue for three decades, growth flat. Unit economics 7.0: pure subscription at $19.99 per year plus a $90 data add-on, never raised a dollar, still fifteen people. Team 5.5: one technical founder with thirty years on the problem, a small stable team, no second leader, no ambition beyond the current shape. Founder trust 6.5: consistent public history, pricing disclosed in every App Store listing, no adverse findings. Mean of 4.0, 6.5, 6.5, 7.0, 5.5, 6.5 is 36.0 divided by 6 equals 6.00. Adjusted down 0.25 because the only vertical where the database could have converted into resale economics, trading cards, is already taken by a free competitor with nearly three times the ratings. Base 5.75.

The formula 5.75 + 1.00 (EV, $0 raised per Harmonic) + 0.25 (Netherlands, EU) + 0.00 (chain) = 7.00.

For Daxos, price and verdict DAXOS 1.0. Thirty years without a round is not a data gap, it is the company's position. There is no round to price, no founder seeking a partner, and no path by which a $100k to $500k check into a fifteen-person Amsterdam lifestyle business returns a seed fund. CLZ Core would suit a larger collectibles platform, making this a roll-up target, not a venture investment. PASS. Keep the card as the honest ceiling marker for what thirty years of excellent execution in this lane produces.

What could not be checked Revenue, subscriber count, trial conversion, CovrPrice attach rate, Android install base, and anything behind clz.com, which returned 403 on every fetch. The 1996 versus 2000 founding discrepancy between the company history and Harmonic was not resolved.

back to the ranked table

The strongest execution in the tool lane, three people and no recorded funding, with a sourcing wedge that the platforms do not give away free and a real $25 per month price already being paid.

What it is

Flippr scans secondhand goods for resale value. Point the camera at one item or a whole shelf, and it identifies every item in the frame, pulls eBay sold comps and Amazon pricing, shows sell-through rate and a "Flippr Score," and then autolists to eBay in one tap. Physical media in bulk is the core use case: movies, games, books, CDs, vinyl, scanned from cover, spine, cartridge or disc (useflippr.com and App Store listing, both 2026-09-16).

The important distinction from everything else in this batch: the primary job is sourcing, not listing. Flippr answers "should I buy this shelf for $200" while standing in a thrift store. Listing is the second act.

Founders and team

Collin McGregor, co-founder, from 2024-08. Brigham Young University, physics with machine learning and math, and applied mathematics with machine learning, 2019 to 2025. Levinthal Fellow at Sandbox, BYU's startup program, 2024-01 to 2025-04. Palantir American Tech Fellowship, 2025-06 to 2025-07. Research assistant in the BYU Biophysics Lab 2023 to 2024, plus machine learning and software engineering internships at Nexu, UI Charitable and Village Capital. One prior startup, Portol, co-founded 2024-07 and left 2024-12, which is a five-month attempt rather than a failure of scale. Now based in Los Angeles (Harmonic person 163025597).

Technical founder confirmed: ML coursework, research posts, engineering internships, and a shipped computer vision product on three surfaces.

Braydon Anderson joined as Investor and Advisor in 2025-10. Utah Valley University marketing, four years at Qualtrics through 2016, then VP of Operations at Clozd from 2017 to 2025, now CRO at Interval (Harmonic person 25070320). That is a credible go-to-market advisor and the "Investor" half of his title implies angel money that Harmonic has not recorded.

Harmonic reports headcount 3 and no other named staff. No CTO exists separately from the founder.

Checks run: Harmonic company record and both Harmonic person records, company site including pricing, FAQ, terms and creator program pages, App Store lookup for id 6746770511, LinkedIn and Facebook company links from Harmonic socials. No adverse record found in the checks run.

Product and traction

iOS app "Flippr: Find, List, Earn," id 6746770511, seller Flippr, Inc., released 2025-07-17, version 3.5.37, last updated 2026-09-10, 1,570 ratings at a 4.75 average as of 2026-09-16 (itunes.apple.com/lookup, 2026-09-16). Android is live and a web app exists. Harmonic web traffic 36,240. The site claims "over 60,000 resellers," unverifiable.

Version 3.5.37 after fourteen months is a fast release cadence for three people.

Set that in the lane. Reclaim launched within a week of Flippr and has 86 ratings. Hero, a year older, has 2,060. alias, five years older, has 6,848. Flippr is the fastest ratings accumulation of any app in this batch.

Supporting surfaces: a Reseller Academy, an item value index, a compare-apps page, a blocked buyer list, a creator affiliate program with published tiers and guidelines, and a business API. That is a company building a funnel, not a weekend app.

Business model and unit economics

Flippr Pro is $25 per month or $95 per year, capped at 3,000 scans per month (useflippr.com/upgrade, 2026-09-16). A hard scan cap is the correct design when the marginal cost is vision inference plus marketplace API calls, and it gives an obvious upsell ladder.

Two things make the economics better than the rest of this lane. First, the customer is an active flipper who uses the tool before every buying decision, not a declutterer clearing a garage once. Retention should follow usage. Second, the value is measurable in dollars per sourcing trip, which supports a $25 price.

The eBay integration goes through the official API, with the standard disclaimer in the footer. That is meaningfully safer than browser automation into Facebook Marketplace, though it still leaves Flippr exposed to eBay's February 2026 user agreement language on agentic bots if the autolister crosses from API into agent behaviour.

Position in this market

The category brief's kill shot is that every leg of the consumer scan-value-list loop already ships free inside the platforms. That is true of listing. It is not true of sourcing. eBay does not stand in a Goodwill aisle and tell you which forty of these two hundred DVDs are worth buying, and neither does Google Lens at bulk-frame scale. Flippr also sits where the profitable survivors sit, bootstrapped with three people, rather than where the graveyard sits.

The ceiling still binds. The addressable market is serious resellers, maybe low hundreds of thousands of people in the US, at roughly $100 to $300 a year. Even excellent execution lands a business worth tens of millions, not a fund returner. ScoutIQ and Keepa occupy the adjacent Amazon version of this and never became venture outcomes.

Base decomposition

Market and competition 4.0: small but real market, and the sourcing job is not one the platforms give away. Product and wedge 5.5: bulk multi-item recognition from a single shelf photo, on iOS, Android and web, is the most differentiated product in this batch. Traction 5.5: 1,570 ratings in fourteen months, best in the sample, plus 36,240 monthly web visits. Unit economics 5.0: real $25 price already being paid, scan cap controlling COGS, a usage-driven customer. Team 4.0: capable solo technical founder with ML training, but three people and no prior exit. Founder trust and disclosure 4.5: honest eBay API disclaimer, published refund policy, review claims stated conservatively below the real number.

Mean of 4.0, 5.5, 5.5, 5.0, 4.0, 4.5 is 4.75. Adjusted up 0.25 because the sourcing wedge sits outside the free-platform kill zone and paid conversion is already priced and live. Base 5.00.

Formula

5.00 base + 1.00 EV + 1.00 geo + 0.00 chain = 7.00 company.

For Daxos, price and verdict

Harmonic shows no round and currently_raising is null, and the advisor's investor title suggests whatever money exists came from one angel. A three-person team with a live subscription and no venture overhang may not want a seed round at all, and may be right not to. If a priced round appears in the low single-digit millions post, this is the one company in the batch where a $100k to $500k check is defensible, on the sourcing wedge and on the founder. The honest limit is that the category ceiling makes a fund-returning outcome unlikely even if everything goes right, so the check would be a good-business bet, not a power-law bet. Daxos adds little here beyond introductions. Daxos fit 5.0. MEET, and ask what revenue is and whether he is raising.

What could not be checked

Revenue, paying-subscriber count, the 60,000 user claim, churn, whether the autolister uses eBay's API end to end, the size and terms of the 2025-10 angel money, and Android install count.

back to the ranked table

Best consumer traction in the whole scan-and-list set at 39,656 iOS ratings, but it is a churny consumer subscription in the tool lane and YC F25 pricing puts it out of Daxos range.

What it is

A consumer iOS app that turns an image, barcode or text prompt into a listing and pushes it to several marketplaces at once. Harmonic external_description, pulled 2026-09-16. The site adds inventory tracking, gross inventory value, buyer messaging across channels, USPS label creation and door pickup. Claimed integrations: eBay, Depop, Facebook, Mercari, Goat (sellraze.com, 2026-09-16). This is the exact consumer scan-value-list loop the category brief covers, executed better than anyone else in the set.

Founders and team

Jeff Mao, Founder, Georgia Tech. Ran Philosophy Robotics from 2017 to 2022-07, which he describes himself as founded in high school "to build sneaker bots and retail automation tools" with "advanced scraping and botting systems" (Harmonic person 108142438, experience description). Software engineering intern at LexisNexis Risk Solutions, summers 2020 and 2021. Z Fellows. Technical by any reading.

Tyler Ma, Cofounder and CTO since 2022-08, earlier "Cofounder, Backend Developer" from 2022-05. University of Georgia and Georgia Tech. His own description: built the backend against REST APIs of ecommerce platforms including Amazon Selling Partner, and won the Georgia Tech Create-X startup competition against 40 teams (Harmonic person 72595780). Research work at the UGA Innovation Factory on real-time factory dashboards. Technical.

Carter Cote, Co-Founder 2024-03 to 2024-12, then left. His own record claims SellRaze went "0 -> 300K users. $1M+ARR" on his watch, and that he left to build Phia, where he was Chief Product Officer from 2025-01 to 2026-02 and claims 0 to 1M users in 8 months and a $35M Series A at $180M (Harmonic person 113252543). Now founder of Ascend Studio since 2026-04. A co-founder leaving after nine months matters; that he left for a competitor in adjacent shopping AI matters more.

Checks run: Harmonic company and three person records, company site, App Store listing and description, one general web search on the founders. No litigation, dissolved entity or disputed-credential record surfaced. No adverse record found in the checks run. I did not run LinkedIn scraping or court records.

Two disclosure flags, neither disqualifying. First, a third-party YC-adjacent summary says SellRaze "has raised over $1m" while Harmonic records $500,000 across two rounds as of 2026-09-16; take Harmonic. Second, the $1M+ ARR figure for 2024 exists only on a departed co-founder's own profile and is not corroborated by the company.

Product and traction

US App Store, id 6455042085, checked 2026-09-16: 39,656 ratings, 4.81 average, released 2023-09-11, version 5.22 shipped 2026-09-11. That rating count is the single best traction number in this competitor set and sits in the same order of magnitude as Collectr's 64,932 in a vertical with liquid comps. Weekly-ish release cadence. Site claims "Join 500,000+ sellers" and a Tech Square Atlanta piece cites 200,000 sellers and 500,000 items listed over eight months (techsquareatl.com, 2025-04-16). Headcount 11 (Harmonic).

Business model and unit economics

Pure consumer subscription with listing caps. Premium $44.99 a year or $3.75 a month equivalent, 25 listings a month. Pro $20 a month, 125 listings. Elite $30 a month, 250 listings. Enterprise on contact (sellraze.com/pricing, 2026-09-16). No take rate on sales. This is the weak leg. A casual seller lists a few things after a closet clean-out and churns; the heavy seller who would pay $30 is the pro reseller already served by the bootstrapped cross-listers. Every scan carries model inference cost against a $3.75 monthly floor. And the same identify-price-list function now ships free inside eBay, Facebook Marketplace, Depop, Mercari and Poshmark, plus free at the OS layer.

Position in this market

The best consumer execution in the tool lane, which is still the tool lane. Platform dependence is the live risk: there has never been a public Facebook Marketplace C2C listing API, Mercari's 2024-08 change broke cross-listers silently, Poshmark suspended cross-lister users twice in 2025, and eBay's 2026-02 user agreement bans LLM-driven agentic bots. SellRaze claims one-click Facebook integration, which implies automation rather than a sanctioned API. A CEO whose prior company built retail bots is unusually well equipped to keep that working and unusually exposed if a platform decides to make an example.

Base decomposition

Market and competition 3.0: large TAM, but the tool lane has produced no venture outcome and the incumbents now give the product away. Product and wedge 5.5: fastest and most complete consumer loop in the set, extended past listing into pricing, authentication, shipping and pickup. Traction 6.5: 39,656 ratings at 4.81 is category-leading and independently checkable. Unit economics 3.5: capped consumer subscription from $3.75 a month, churny base, inference cost per scan, no take rate. Team 5.5: two technical founders, five years on the same problem, YC F25, real shipping cadence; one co-founder departure. Founder trust and disclosure 4.0: candid about the bot origins, but a funding figure inflated elsewhere and an uncorroborated ARR claim in circulation.

Mean = (3.0 + 5.5 + 6.5 + 3.5 + 5.5 + 4.0) / 6 = 28.0 / 6 = 4.67. Adjust +0.25 because the rating count is hard evidence of consumer pull that no peer in this set can show. Base = 4.92.

The formula line

COMPANY = 4.92 Base + 1.00 EV + 1.00 Geo + 0.00 Chain = 6.92

For Daxos, price and verdict

Harmonic estimates a $2.0M post on the $500k pre-seed, flagged ESTIMATED from typical 25 percent dilution, so it is an inference not a term sheet. That price is not the live price. This is YC F25, meaning they are raising or about to raise on post-Demo-Day terms, and a company with 39,656 ratings coming out of YC will clear $15M to $25M post. A $100k to $500k check at that price buys 1 to 3 percent of a business whose category ceiling is a mid-eight-figure tools acquisition. That does not return a seed fund.

At or under a $6M post the math would work and this would be an IN conversation. At YC seed pricing it is a PASS on price, not on quality. Daxos adds nothing here that YC does not already supply.

VERDICT: MEET. Take the founder call. They are the sharpest operators in this competitor set and the conversation is worth more than the check, both as a read on where the category is heading and because Jeff Mao is someone to know for a later company. Do not chase the round at YC pricing.

What could not be checked

Revenue, paid conversion and subscriber count: no source. Churn: no source. Google Play presence and installs: not confirmed from the site. Whether marketplace integrations run on official APIs or browser automation: not disclosed. Whether the 2025-09 round is the full YC F25 raise or a first tranche. Any current fundraise terms.

back to the ranked table

The comps layer for the watch vertical, bootstrapped to seven years and four people with zero dilution, and almost certainly not buyable.

What it is

WatchCharts is the price index and valuation layer for pre-owned watches. It tracks open market prices, publishes market indexes, runs a price guide, and sells a Premium subscription to collectors and dealers. It has extended into WatchCharts Marketplace, where users list watches, make and accept offers and leave reviews, with listings carrying pricing labels from its own historical data (watchcharts.com/subscribe, /watches/about and marketplace.watchcharts.com/about, per search results read 2026-09-16; the site sits behind Cloudflare and challenged every fetch).

Founders and team

Charles Tian, founder. An engineer at IBM, he started buying watches and kept a spreadsheet of what he paid and what things were worth. That spreadsheet, in his own account, "snowballed into the company that we are today" (ablogtowatch.com interview, read 2026-09-16). He writes the market updates himself.

Technical: yes, he built the data platform. No cofounder is visible in any source. Headcount 4 after seven years, the bootstrapped survivor profile rather than a venture trajectory. Note the App Store publisher is "WatchEnthusiasts LLC," not a C-corp (iTunes lookup, 2026-09-16), and a seed fund cannot take preferred stock in an LLC without a conversion.

Checks run: Harmonic enrich, iTunes search and lookup, web search on the company, the founder and a possible Hodinkee acquisition, aBlogtoWatch and Worn & Wound interviews, Robb Report, ZoomInfo and RocketReach, attempted site fetch. No adverse record found in the checks run: no litigation, no dissolved entities, no disputed claims, no inflated titles, no acquisition.

Product and traction

Self-reported: over 500,000 collectors, and price tracking across 29,871 watches on the secondary market (watchcharts.com pages, via search results read 2026-09-16). The indexes are described as trusted by watch brands and financial institutions, and Robb Report lists WatchCharts among three market data tools serious collectors should use.

iOS: 1,330 ratings at 4.81, live since 2020-11-18, last updated 2026-06-24. Small by rating count, but the app is a companion to the web product, which is the real surface.

Business model and unit economics

Subscription, free tier plus paid Premium for collectors and dealers, plus a marketplace that monetises transactions. The best unit economics in the batch by a wide margin, and the evidence is the cap table: seven years, $0 raised, four people, still shipping. A company that survives that long with no outside money covers its costs from revenue. No burn to fund, no preference stack. The ceiling is equally clear: subscription data for watch collectors is a small addressable base, and a declining secondary market shrinks subscription appetite and marketplace GMV together.

Position in this market

This is the vertical-with-liquid-comps pattern, the watch equivalent of PriceCharting for games or Card Ladder for cards. The data asset is hard to rebuild: seven years of price history across nearly 30,000 references is not something a funded competitor buys back with a round. The marketplace extension is the correct and harder move, converting a reference product into a demand network.

Not a seller tool, no platform automation dependence, no overlap with a cross-lister. Threat to Moe: 1. The real risk is Chrono24, which sits on far more transaction data and could ship a comps product whenever it likes.

Base decomposition

• Market and competition: 4.0. Underserved data niche, weak direct competition, but a small subscription base and Chrono24 sitting on better raw data.
• Product and wedge: 6.0. A 29,871-reference index with seven years of history and institutional users is compounding and hard to replicate.
• Traction: 5.0. 500,000 claimed collectors, 1,330 iOS ratings at 4.81 since 2020. Claims self-reported.
• Unit economics: 6.0. Seven years, four people, zero dilution. Revenue covering cost, demonstrated not projected.
• Team: 4.5. Technical, obsessive founder seven years in, but four people and no second executive.
• Founder trust: 4.0. Public index methodology and bylined market updates. User numbers unaudited, no financials.

Mean = (4.0 + 6.0 + 5.0 + 6.0 + 4.5 + 4.0) / 6 = 29.5 / 6 = 4.92. No adjustment; the data asset's strength and the ceiling's weakness cancel. Base = 4.92.

The formula

4.92 + 1.00 (EV, $0 raised, the $0 to $1M tier) + 1.00 (Geo, USA) + 0.00 (Chain) = 6.92.

For Daxos, price and verdict

The highest company score in this batch, and the least likely check. Seven years bootstrapped, $0 raised, four employees, an LLC wrapper and no sign of a round means there is probably nothing to buy. If Tian did raise he would price off revenue, and the ceiling for a watch comps subscription is a strategic sale in the tens of millions, which does not return a fund on a $100k to $500k check.

MEET, not because a deal is available but because the conversation is worth more than the deal. Tian holds the best proprietary comps dataset in the watch vertical, and liquid comps are the one durable asset in resale. A call answers whether he intends to raise, how much revenue is subscription versus marketplace, and whether the marketplace converts the index audience into transactions. DAXOS fit 4.0.

What could not be checked

Revenue, subscriber count, pricing tiers, marketplace GMV and take rate, profitability. The site could not be read directly: Cloudflare challenged curl, the jina reader and WebFetch, so product detail comes from search snippets of WatchCharts' own pages. Corporate structure beyond the LLC name. Non-public cofounders. Google Play. No LinkedIn scrape: the Apify actor errored twice and was stopped.

back to the ranked table

Shopify-for-secondhand-stores built by two people who ran one, with the best revenue model in this batch and almost no public evidence of scale.

What it is

Ribbn is an operating system for secondhand and consignment stores: inventory management with AI-assisted listing, a webshop builder for one-of-a-kind items, a mobile point of sale on Stripe and Zettle, reusable RFID tags for locating stock, and seller sourcing with commission tracking and self-serve payouts (ribbn.ai, 2026-09-16). It also publishes white-label iOS apps for customer stores. QuickList claims 99% accuracy from a few photos and a 70% cut in entry time.

Founders and team

Vendela Ragnarsson is co-founder and CEO. Joshua Beach is co-founder and CTO. Per the company story page (ribbn.ai/our-story, fetched 2026-09-16) they built a mobile app in 2019 to run their own Stockholm secondhand store, ReRobe. In 2021 Ragnarsson pitched the technology on Draknästet, the Swedish Shark Tank, producing inbound interest from other resale operators wanting to license it. In 2022 the platform "fully powers its operations, generating more than $500K in revenue" at ReRobe. In 2023 they launched Ribbn. Leadership also includes Kaja Howard (CDO) and Viktoria Mahrberg (PM). A co-founder CTO on the same codebase since 2019 is the strongest technical signal in this batch.

Checks run: Harmonic enrich and person_relationships (zero people), company story and pricing pages, App Store developer record, web search on the company and on Joshua Beach, Crunchbase profile. No adverse record found in the checks run. The New York address is a PMB mail drop for a Stockholm-rooted team, unremarkable for a Delaware-wrapped European startup but worth noting.

Product and traction

Public traction evidence is thin. The flagship app has 3 ratings. Three white-label store apps are live with zero ratings each, which tells me at least three paying stores shipped branded apps and nothing more. Harmonic headcount is 6. The site names no customer count, no store count and no revenue. The only hard number anywhere is the founders' own store doing more than $500k in 2022. Low ratings are expected for a B2B product, but it leaves no independent traction proxy at all.

Business model and unit economics

Published pricing (ribbn.ai/pricing, 2026-09-16): Starter $70 per month plus 1.5% on in-store sales, 2 users. Business $295, 1.5% in-store and 0% online, 4 users, adds QuickList and analytics. Commercial $595, 1% in-store and 0% online, 6 users, adds digital consignment, seller self-payouts, web POS with a terminal and an RFID gun. Enterprise custom.

Best model in the batch: a real SaaS price against a business budget plus a take on gross merchandise value, embedded in point of sale and RFID hardware. Once a store's tills, tags and consignor payouts run through Ribbn, switching means retraining staff and re-tagging inventory.

Position in this market

Ribbn is not in the consumer scan-value-list lane and is not exposed to the compression the brief documents. Free AI listing inside eBay, Facebook Marketplace, Depop, Mercari and Poshmark does not touch a store's floor operations, tills or consignor accounting. Platform dependence does not bite either: the missing Facebook Marketplace C2C API, Mercari's August 2024 break and Poshmark's 2025 suspensions hit crosslisters, not a POS vendor. Competition is Shopify, SimpleConsign, ConsignPro and Ricochet.

Base decomposition

• Market and competition 4: real and underserved, but the count of stores able to pay $295 per month is low tens of thousands and small retail churns hard.
• Product and wedge 5: broad and workflow-true. POS, RFID tags, consignment payouts and a resale-native webshop are not a weekend clone.
• Traction 3: 6 headcount, 3 ratings on the main app, three white-label stores live, no disclosed customer count or revenue.
• Unit economics 6: $70 to $595 per month plus 1 to 1.5% of in-store GMV, sold to a business, with POS and hardware switching costs. Best in the batch.
• Team 5: operator-founders who ran the store the product serves, a named co-founder CTO, six years on one problem. The TV pitch producing inbound licensing demand is real pull.
• Founder trust and disclosure 5: dated public story, named leadership, published pricing including take rates, funding consistent with Harmonic. Only soft spot is the PMB address.

Mean (4+5+3+6+5+5)/6 = 4.67. No adjustment; product and economics strength offsets the traction weakness. Base 4.67.

Formula

4.67 + 1.00 + 1.00 + 0.00 = 6.67

For Daxos

DAXOS 5.5. Ribbn is the best-built company in this batch and the only one whose revenue line would survive a diligence spreadsheet. It last raised 2025-06-20 with $886k total, so a 2026 round is plausible at single-digit millions post, fair for this quality.

The problem is the ceiling. Three thousand stores at $295 per month is about $10.6M of ARR: a very good outcome for two founders and a poor one for a seed fund, and the brief is explicit that only demand networks have produced venture-scale outcomes here. The upside case requires the 1 to 1.5% GMV line to grow into a payments and consignment-network business across a connected store base, which Ribbn's "tap into a thriving network" copy gestures at but does not evidence.

MEET. The questions are narrow: how many paying stores, what is net revenue retention, and what share of revenue already comes from the GMV take rather than the subscription. If the GMV line grows faster than the seat line, this becomes a real conversation about $100k to $250k.

VERDICT MEET.

What could not be checked

Paying store count, revenue, retention, churn, the basis for the 99% QuickList accuracy claim, the Sweden versus US split, Google Play presence, and whether a new round is being raised.

back to the ranked table

AI pricing for the thrift store floor, sold to Goodwill and Planet Aid at $249 a store, with the only customer evidence in this batch that a third party put on the record.

What it is

Thriftly prices, tags and lists inventory for thrift stores. A worker photographs an item, the system identifies brand, category and size, sets a price from brand and store location, prints a tag in seconds and can route the item online with an AI-written description. Harmonic says live in 70+ stores; site copy claims 15% more revenue on the same labour (thriftly.com, 2026-09-16). Value is created on the store floor, on items that mostly never touch a marketplace. The competitor is a worker guessing a price, not eBay.

Founders and team

Andy Downard, co-founder and CEO, was the first hire on the product team at Happy Returns, the returns-logistics company bought by PayPal and later by UPS, and a senior product manager at Shopify. Josh Packard, co-founder. Their prior reverse-logistics startup was acquired (BusinessDen, 2025-02-10; Google public policy story, both accessed 2026-09-16). Press describes Downard as having built the tagging and description system himself, so I record a technical founder, though no GitHub confirms it. Domain fit is exact.

Checks run: Harmonic enrich and person_relationships (zero people), company product and pricing pages, web search on both founder names, BusinessDen, CBS Colorado, 9News, Retail TouchPoints, ZoomInfo, RocketReach, iTunes search. No adverse record found in the checks run.

Product and traction

Best-evidenced traction in the batch, and most of it comes from outside the company. BusinessDen reported in February 2025 that the technology ran loyalty programs for 62 Goodwill stores in Michigan and Florida, and that Thriftly opened its own 1,000 square foot Denver thrift store to test the product on itself. The Big Reuse case study names a Brooklyn nonprofit, quotes executive director Justin Green on record, and claims clothing revenue up 15% in month one and 26.5% inside a year. Retail TouchPoints, CBS Colorado and 9News have covered it. Site logos: Goodwill, Planet Aid, St. Vincent de Paul. Caveat: counts and lift are company-reported, and the homepage dashboard is labelled sample data.

Business model and unit economics

Published pricing (thriftly.com/pricing, 2026-09-16): Basic Essential $99 per store per month for tagging, donor CRM and loyalty. Pro AI-Powered $249 adds the pricing engine, brand detection, photo capture and online listing, and sell-through analytics. Enterprise for 20+ stores at typically $149 to $199. At 70 stores and a blended $150 to $249, run-rate revenue is roughly $126k to $209k. Small, but per-store pricing against a chain is the right shape: one Goodwill region is dozens of stores on one contract.

Position in this market

Thriftly is outside the lane that kills companies in this category. It depends on no marketplace API, so Mercari's August 2024 change, the Poshmark suspensions of 2025 and eBay's February 2026 agentic-bot ban do not reach it. Nor can the platforms give this away free: they ship photo-to-listing with suggested prices for consumers, but none operate a store floor, print tags or run a donor CRM. The counterweight is the customer: nonprofit chains buy slowly and on constrained budgets, which is why the price is $249 and not $2,490.

Base decomposition

• Market and competition 4: real, but the paying universe is tens of thousands of stores at a nonprofit-constrained price, so the ceiling is modest.
• Product and wedge 6: pricing the floor, not the marketplace. No platform dependence, no terms-of-service risk, a measurable ROI claim, a tag printer in the workflow.
• Traction 5: 70+ stores per Harmonic, named chains, a customer with a quoted director and a year-one figure, press in four outlets.
• Unit economics 5: $99 to $249 per store per month, land and expand across chains, zero added labour. Capped by nonprofit budgets and a slow sale.
• Team 6: an early Happy Returns product hire and Shopify PM plus a co-founder from an acquired reverse-logistics startup, dogfooding in their own store.
• Founder trust and disclosure 5: named customers on record, press, published pricing, Harmonic-consistent funding.

Mean (4+6+5+5+6+5)/6 = 5.17. Adjust -0.25 for 28 months since the last round with only 4 people, which usually means growth did not clear a seed bar. Base 4.92.

Formula

4.92 + 0.75 + 1.00 + 0.00 = 6.67

For Daxos

DAXOS 6.0. The one company in the batch where a check is arguable rather than academic. Total raised is $1.47M at pre-seed, last priced 2024-05-06, so a seed at a high single-digit post is the plausible ask and $100k to $250k fits. Daxos would be buying exact domain history, a wedge the platforms structurally cannot commoditise, and customers who go on the record about ROI.

Two things hold it back from IN. First, 28 months without a new round on a four-person team is a signal, and the likely explanation is that store adds have been slower than the press suggests. Second, the ceiling: charity retail software at $249 a store does not return a seed fund. The fund-returning version requires moving from pricing into the transaction, taking a cut of the online listings it already generates, and there is no evidence of that yet.

MEET, leaning to a check. Ask for store count month by month since 2024, retention inside the Goodwill and Planet Aid accounts, and what share of priced items ends up listed online. If store count compounds and a transaction take is credible, this moves to IN under roughly $12M post.

VERDICT MEET.

What could not be checked

Current store count against the 70+ figure, ARR, churn and retention, whether the Goodwill relationships are pilots or signed chain contracts, whether a round is being raised, and Shopify app install counts.

back to the ranked table

The only company in this competitor set sitting on the demand side rather than the tool side, run by the strongest founder in the set, and deliberately confined to a few Atlanta neighborhoods after two years.

What it is

Not a seller tool. Loop buys the item. Photograph a household object, get a guaranteed cash offer in about eight seconds, accept, leave it on the porch, a driver collects it the next weekday, and Loop pays by Venmo within seven days whether or not the item resells (loopstuff.com, read 2026-09-16). Goods go into a local buy feed the company says holds 10,000 or more items, with seven-day free returns.

Loop takes principal risk and owns the inventory. It is a managed local recommerce network with an AI pricing model attached, not a cross-lister and not a valuation app.

Scope is narrow by design: household items one person can carry, working, worth more than $20. No clothes, shoes, art, collectibles or large furniture. Service area is intown Atlanta, Decatur first for about a year, then Brookhaven (Hypepotamus, 2026-02-04).

Founders and team

Seth Radman, Co-Founder, CEO and CTO since 2024-09. Georgia Tech 2012 to 2017, engineering track. Before founding anything: structural test, stress and mechanical systems engineer at Honda Aircraft Company across 2014 to 2016, then product development engineer at Siemens. Four companies, two acquisitions: Plutonium Apps 2014 to 2019, Crescendo 2016 to 2020 acquired by Ultimate Guitar, Upbeat 2020 to 2023 acquired by MakeMusic, and Co-Founder and CTO of Infinite Giving 2021 to 2023, which he says raised $4M. Forbes 30 Under 30. Plutonium Apps, LLC is still the App Store seller of record for Loop (LinkedIn via Apify harvestapi, 2026-09-16; iTunes, 2026-09-16).

Technical, holds CTO as well as CEO, and has sold two companies. The strongest founder profile in this competitor set by a wide margin.

Dave Payne, Co-Founder. Co-Founder and CEO of Scoutmob, co-founder of Switchyards, Managing Director of Techstars Atlanta, and since 2022-07 Co-Founder and Managing Partner of Neighborhood Studios, the studio Loop spun out of (Techstars blog; Hypepotamus, 2026-02-04). Scoutmob was one of Atlanta's best known consumer startups and no large exit for it appears on the public record. His contribution is distribution and the studio, not engineering.

Checks run: Harmonic company record, which returned zero person relationships and no headcount; the site and FAQ; App Store listing and description; Google search on both founders; two Hypepotamus pieces; the Techstars profile; an Apify LinkedIn pull. The LinkedIn pull returned Radman correctly and returned the wrong Dave Payne, a public relations executive in Orlando, so Payne's record rests on press and studio material. That is a stated limit, not a finding. No adverse record found in the checks run on either founder.

Product and traction

112 ratings at 4.54, released 2024-08-13, updated 2026-07-26 v1.81 (iTunes, 2026-09-16). Instagram 863 followers (Harmonic, 2026-09-16). Version 1.81 in two years is a team that ships.

112 ratings over twenty-five months is thin, and should be read against a deliberately hyperlocal strategy. The company's unverified claims are larger: 1,200 or more reviews, 10,000 or more items in the feed, thousands of items picked up with none stolen. The 1,200 figure does not reconcile with 112 App Store ratings and no source is given.

One quality note: the live site still carries unedited Webflow template copy, including a section about mobile vans selling snacks and a line beginning "At Scribbler, we specialize in creating unique, eye-catching designs." Small, but it sits on the homepage of a company asking strangers to leave valuables on a porch.

Business model and unit economics

Loop buys low, resells higher, eats the spread. No take rate, because there is no marketplace fee: margin is the gap between the guaranteed offer its model generates and what the item clears for locally, less pickup, warehousing, quality check, returns and payment costs. Buying regardless of resale is the whole promise and the whole risk.

Controllable, because Loop sets the buy price, and capital-hungry, because every purchase is working capital in a warehouse. The $20 floor and the clothing exclusion show they already know which items cannot carry the pickup cost. Margin, volume and sell-through are undisclosed.

Position in this market

The only venture-scale outcomes here have been demand networks, and seller tools have never produced one. Loop is the only company in this set on the demand side, and the only one where the seller's real objection, the meetup and the haggling and the not-knowing-if-it-will-sell, is removed rather than automated.

The counterweight is that this is local operations. Each metro needs drivers, a warehouse and inventory capital, so it compounds slowly and expensively. Two neighborhoods in two years is the honest read on that pace.

Base decomposition

• Market and competition 4.5: demand side rather than tool lane, the one structural position in this category that has produced an outcome, discounted because hyperlocal ops caps the prize per unit of capital.
• Product and wedge 5.0: guaranteed instant offer, porch pickup, payment whether or not it sells. Nothing else in this set removes the actual reason people keep clutter.
• Traction 2.5: 112 ratings at 4.54 over twenty-five months, 863 Instagram followers, two neighborhoods, larger self-reported numbers unverified and one inconsistent.
• Unit economics 3.0: principal risk on every unit, working capital in inventory, driver and warehouse cost against $50 items, margin undisclosed, offset by full control of the buy price.
• Team 6.5: a technical CEO and CTO with two acquisitions and aircraft engineering behind him, paired with an Atlanta hyperlocal operator who runs the studio.
• Founder trust and disclosure 4.0: backgrounds verify cleanly, but the site ships unedited template copy and an unsourced 1,200 reviews claim.

Mean = (4.5 + 5.0 + 2.5 + 3.0 + 6.5 + 4.0) / 6 = 25.5 / 6 = 4.25. Adjust +0.25 because demand-side position plus a twice-exited technical founder is the only combination in this set matching the shape the category has rewarded. Base = 4.50.

Formula

4.50 + 1.00 (EV, Harmonic funding_total $0, $0 to $1M band) + 1.00 (Geo, USA) + 0.00 (Chain) = 6.50.

The EV term has not earned its place here. A studio spinout running drivers and a warehouse for two years is not running on nothing; Harmonic simply has no record of the studio's capital. Treat 6.50 as formula output and 4.50 as the business.

For Daxos, price and verdict

Harmonic shows no round, no valuation, no valuation_info, and there is no public evidence Loop is raising. The problem for Daxos is the cap table, not the company: a studio spinout usually leaves the studio holding a large founding stake, compressing what a $100k to $500k check owns, and the next real financing is an ops and inventory expansion round rather than a seed. A check this size cannot return a fund unless Loop becomes a multi-metro network, which would take far more capital than its pace suggests it is pursuing.

Against that, Radman is the one founder in this set worth knowing regardless of this company, and Loop is the one model pointed the right way. DAXOS = 5.5. VERDICT MEET. Worth a conversation to learn buy-price margin, sell-through and cost per pickup, and to understand the studio's ownership, before any view on price. Not a check on current information.

Threat to Moe: 1. Different model, different side of the market, single metro, no overlap in the tool lane.

What could not be checked

Items bought and sold, gross margin per item, sell-through, cost per pickup, inventory on hand, revenue. Whether Neighborhood Studios put capital in and how much it holds. Whether Loop is raising. The 1,200 reviews and 10,000 items claims. Dave Payne's LinkedIn record, since the scrape returned a different person and the profile cap was reached. Google Play presence.

back to the ranked table

A structured-catalogue vertical marketplace for Warhammer miniatures, which is the one shape this category has ever rewarded, filling real supply at eleven months old.

What it is

Miniswap is a marketplace for tabletop wargaming miniatures, starting with Warhammer. General marketplaces fail this hobby: filters do not map to armies and factions, free-form listings mean products are constantly misidentified, and counterfeit recasts crowd search. Miniswap's answer is a catalogue of more than 20,000 unique miniatures that sellers match their listings against, so every listing inherits structure rather than inventing it (ycombinator.com/companies/miniswap, read 2026-09-16). YC Fall 2025.

Founders and team

Zak Singh, co-founder and CEO. Former Cambridge CS PhD, dropped out to build this. East Bay native, has played Warhammer for two decades (YC bio; techfundingnews.com, read 2026-09-16). Technical and the domain expert, the right pairing of the two. Will Hanna, co-founder and COO. Former strategy consultant, dropped out of law school for the company. Canadian. The pair met at Cambridge (startupmag.co.uk, read 2026-09-16).

Two program dropouts is commitment, not a flag. The gap is that neither has run a marketplace, an operations business as much as a software one. Headcount 5.

Checks run: Harmonic enrich, YC company and launch pages, web search on both founders and the funding round, site render, iTunes search. No adverse record found in the checks run: no litigation, no dissolved entities, no disputed claims, no inflated titles. The reported $3.5M seed is consistent with Harmonic's $4.0M across two rounds.

Product and traction

Live and stocked. Listing counts from the homepage on 2026-09-16: Space Marines 1,542, Chaos Space Marines 498, Orks 461, Astra Militarum 441, Aeldari 359, Tyranids 286, Necrons 193, T'au 181, Ad Mech 160, Sororitas 148. Over 4,200 live listings in the top ten 40k factions alone, before Age of Sigmar and the rest of the site. Four weeks after launch the company reported 1,000 signups and over $40k of listed inventory (YC launch post). At eleven months, supply is compounding.

No iOS app; web only, behind Cloudflare. No GMV, take rate, buyer count or repeat rate published.

Business model and unit economics

Marketplace take rate, undisclosed. The concerns are order value and shipping. Miniatures sell for tens of dollars, armies for hundreds, and painted models are fragile and expensive to ship. A high single digit take on a sub-$100 order does not cover much support. The offset: buyers spend thousands a year and repeat constantly, so frequency does the work basket size cannot. Policing recasts is a cost centre, and Games Workshop has been aggressive about its IP in secondary channels.

Position in this market

Miniswap is the only company in this batch built in the shape that has worked: a vertical demand network with a structured catalogue, the architecture of CollX and Collectr in cards and PriceCharting in games. Not a seller tool, so the cross-lister graveyard does not apply, and no platform dependence risk because it is the platform, not a bot driving someone else's. Threat to Moe: 1.

The ceiling question is size. The $20B tabletop figure is the primary market; the Warhammer secondary is a fraction of a $3B-by-2030 line, and eBay runs most of it. "Complex hobbies" as expansion is unproven, and each vertical means rebuilding a 20,000-SKU catalogue by hand.

Base decomposition

• Market and competition: 4.5. Underserved niche and the right shape, but small in absolute terms, with eBay entrenched and Facebook groups free.
• Product and wedge: 6.0. A 20,000-item structured catalogue is slow, defensible work general marketplaces cannot replicate, and it solves misidentification and counterfeits directly.
• Traction: 5.0. Over 4,200 live 40k listings at eleven months. Supply works; demand undisclosed.
• Unit economics: 4.0. Low order value, fragile shipping, no published take rate, offset by high frequency.
• Team: 5.5. Technical CEO with twenty years of domain obsession plus a commercial cofounder. No prior marketplace experience.
• Founder trust: 4.5. Early numbers published and consistent with Harmonic. No GMV or take rate. Nothing adverse.

Mean = (4.5 + 6.0 + 5.0 + 4.0 + 5.5 + 4.5) / 6 = 29.5 / 6 = 4.92. Adjust plus 0.25: it is the only company here built as a demand network rather than a tool, the single variable this category rewards. Base = 5.17.

The formula

5.17 + 0.50 (EV, $4.0M, the $3M to $5M tier) + 0.75 (Geo, Canada per Harmonic's Montreal address) + 0.00 (Chain) = 6.42.

For Daxos, price and verdict

Miniswap closed $3.5M in December 2025 and Harmonic estimates an $18M post, marked ESTIMATED. It is nine months into that money with five people and almost certainly not raising. A $100k to $500k check needs a Series A well above $18M or secondary, and at a $40M-plus A the fund-return math needs Warhammer secondary alone to support a multi-billion outcome, which it does not. Expansion into other complex hobbies would change that, and it is best tested in a conversation.

MEET. Worth a founder call before any A, with three questions: take rate and order value, buyer repeat rate rather than listing count, and what the second vertical is and how long its catalogue took to build. Entry price and the expansion thesis are the whole decision.

What could not be checked

GMV, take rate, order value, buyer count, repeat rate, gross margin. Whether the seed was $3.5M in one round or $4.0M across two as Harmonic implies. Games Workshop's posture toward the platform. Whether the company is US or Canadian for incorporation, since Harmonic's address and location fields disagree. No LinkedIn scrape: the Apify actor errored twice and was stopped.

back to the ranked table

YC X25 AI repricing and multi-marketplace distribution for ticket brokers at a 10 percent take rate, with real API access and liquid comps, but priced at an estimated 30M post in front of Automatiq and SkyBox.

What it is

Sellers connect marketplace accounts or upload barcodes; Ticket Wallet reprices every seat against live demand, broadcasts to eight or more marketplaces (Ticketmaster, StubHub, SeatGeek, TickPick, Vivid Seats, AXS, Gametime), and delivers on sale. Tickets stay with the seller rather than going on consignment (ticketwallet.com, 2026-09-16). Adjacent to Moe, not a competitor: same shape (price it, list it everywhere, reprice it), different asset class.

Founders and team

Truman Gelnovatch, CEO, start 2023-11 per Harmonic. Princeton economics, plus a UVA McIntire summer institute and YC Spring 2025. Background is five years flipping tickets by hand. No prior company on record.

Ryan Clare, CTO, start 2023-06. Princeton computer science with an AI and ML focus. Harmonic experience: Software Engineer Intern at Nasdaq (2021 and 2022), Structured Knowledge Applications (2020), Verifi Media (2020). Technical founder confirmed, though the record is internships only, no full-time engineering role.

They met on the Princeton men's soccer team and have been friends roughly eight years (ycombinator.com/companies/ticket-wallet and workatastartup.com/companies/30472, read 2026-09-16). Neither has a prior exit or failure on record. Headcount 16, including five engineers and an ML engineer hired between Nov 2025 and Apr 2026, plus a CFO since Apr 2025 (Harmonic people list).

Checks run: Harmonic company and person records, company site, YC company and jobs pages, web search on both founder names, Crunchbase free listing. No adverse record found in the checks run.

Product and traction

Self-reported on the YC page (2026-09-16), unaudited: $3M or more of ticket inventory onboarded, 100 or more paying sellers, work started with professional teams, 10 percent charged per ticket sold.

iOS app (id 6462701350 is SellOut; this one is id 6463562809, sellerUrl ticketwallet.com), free, 30 ratings at 4.73, released 2023-10-24, updated 2026-09-08 (iTunes, 2026-09-16). Thirty ratings is negligible, but the real product is a web dashboard for brokers, so the app is a weak proxy. Harmonic web_traffic 3,740. Facebook likes flat at 1 for 365 days.

Integrations run through broker and partner APIs at StubHub, SeatGeek, TickPick and Vivid Seats. That is a genuine structural advantage over consumer cross-listers, which have no public Facebook Marketplace C2C API and get users suspended. The mirror risk: Ticketmaster is hostile to automated inventory tooling and the BOTS Act constrains primary-layer automation.

Business model and unit economics

Ten percent of sale price, charged to the seller, stacked on marketplace seller fees already running 10 to 15 percent. That rate is far above what this category supports (cross-listers charge $10 to $40 a month) and holds only because brokers run a business and can measure incremental yield. The risk is compression: SkyBox, the dominant broker ERP, is owned by Vivid Seats and sells flat monthly; Automatiq is a funded rollup selling the same automation. If Ticket Wallet moves to SaaS pricing the model collapses to the category ceiling. Sixteen heads is a fast burn at this revenue scale.

Position in this market

This is the vertical-with-liquid-comps pattern applied to tickets. Every seat has a public comp set updating by the minute, the condition that made trading card pricing a business. It is the closest thing in this batch to a real repricing business. But the buyer is a professional broker, the incumbent stack (SkyBox, Automatiq, Ticket Evolution) is entrenched and well capitalised, and the ceiling for broker tooling is a trade sale, not a network.

Base decomposition

• Market and competition 4.5: $26B resale market, but broker tooling is crowded and incumbents are owned by marketplaces.
• Product and wedge 5.0: credible wedge, real API access, seller keeps the ticket. Not novel against SkyBox plus Automatiq.
• Traction 4.5: 100 paying sellers and $3M inventory is early and self-reported; 30 ratings and 3,740 visits are thin.
• Unit economics 5.0: 10 percent of GMV is the best model in this comparison set, exposed to compression.
• Team 5.0: one domain founder, one technical founder, YC X25, first time, no exits.
• Founder trust and disclosure 5.0: specific numbers published on the YC page; an unrendered "+0%" stat sits on the marketing site, sloppy rather than misleading.

Mean = (4.5 + 5.0 + 4.5 + 5.0 + 5.0 + 5.0) / 6 = 29.0 / 6 = 4.83. Adjust +0.25 for the two things the rest of this market lacks: sanctioned marketplace APIs and a percentage-of-GMV model. Base = 5.10.

Formula

5.10 + 0.25 (EV, $6.5M in the $5M to $10M band) + 1.00 (Geo, USA) + 0.00 (Chain) = 6.35.

For Daxos, price and verdict

Harmonic shows currently_raising null; the last round closed 2025-08-16. Harmonic's valuation_info of $30M is ESTIMATED from the $6M seed at assumed 20 percent dilution, not reported. A $100k to $500k check near $30M post buys about 1 percent of a broker tooling company whose realistic exit is a trade sale at a revenue multiple. That cannot return a Daxos fund. The round is held by Seven Seven Six, Palm Drive and YC, so there is no allocation and nothing Daxos adds that the cap table lacks. DAXOS = 3.0. VERDICT PASS.

Threat to Moe: 2. Different asset class and a B2B buyer, so no direct overlap, but it is proof the scan-price-list pattern monetises when comps are liquid and the API is sanctioned.

What could not be checked

Revenue, sell-through and net retention; the YC figures are unaudited. Whether integrations are official partner APIs or automation on top of broker accounts. Harmonic returned an empty funding_rounds array, so the split between the ~$500k pre-seed and the $6M seed is inferred from funding_total minus last_funding_total. LinkedIn was not scraped; founder history comes from Harmonic person records plus the YC pages.

back to the ranked table

A profitable eight-person family bootstrap selling multichannel listing software to professional Amazon and Walmart sellers, ten years old, zero raised, structurally uninvestable and not in the consumer resale lane at all.

What it is

B2B multichannel listing and inventory software for established sellers. The pitch (listingmirror.com, 2026-09-16) is list once, sell everywhere: create a listing, push it to Amazon, Walmart, eBay and Shopify, sync inventory to prevent overselling, forecast stock, manage warehouses, route orders to FBA or 3PL.

The integration list (integrations page, 2026-09-16) is Amazon including Vendor Central, eBay, Walmart, Newegg, Etsy, Wayfair, Reverb, Wish, Overstock, Temu, SHEIN. No Poshmark, Depop, Mercari, Vinted or Facebook Marketplace C2C. Every connection runs through an official merchant API, which is why this company carries none of the platform-dependence risk that defines the consumer cross-listers, and also why it has nothing to do with somebody photographing a thrifted jacket.

Founders and team

Brian Wawok, co-founder and CEO since May 2017. Master's in Computer Science from DePaul, undergraduate at Illinois Urbana-Champaign (Harmonic person record and employee_highlights, 2026-09-16). A decade of engineering first: Backstop Solutions, software engineer III at CME Group (2009-2013), Sonatype, CTO at eSale Solutions (2015), developer at DataStax (2016-2017). Deeply technical, and the eSale stint means he had touched ecommerce software before founding.

Erin Wawok, co-founder, May 2016 to May 2025. DePaul. Interaction and product design; the about page credits her with design and customer experience. Her Harmonic record shows the role ending May 2025 with no successor role, so after nine years one of two founders has stepped back.

They are married, and the about page states a preference for staying small and answering support calls personally. Harmonic shows 8 employees: two designers, two developers, four support and onboarding.

Checks run: Harmonic person records for both founders with dated role history and education, employee_highlights, the about page, and the Harmonic contact block (primary exec email brian@listingmirror.com). No adverse record found in the checks run. LinkedIn scraping was not run; Harmonic already carried complete histories.

Product and traction

Self-reported homepage metrics (2026-09-16): $175M monthly GMV, 5.7M orders managed monthly, 3.5 years average customer tenure described as twice the industry average. The company's own numbers, no third-party source, so directional only.

Pricing is not published. The Pricing nav link resolves to a feature page about repricing rules, and /pricing-plans and /plans both 404 (2026-09-16). Every call to action is Schedule a Demo or Schedule a Call, with a free trial.

No iOS or Google Play app, so the rating-count proxy does not apply here. Traction evidence is tenure and headcount.

Business model and unit economics

Monthly SaaS quoted by sales, with paid onboarding and US-based support as the stated differentiator. Eight people, ten years, zero outside capital: that arithmetic only works if the business has been cash-generative for most of its life, which is the strongest fact in the file. The cost structure is also the ceiling, since US support staff scale with accounts. A good living, not a venture asset.

Position in this market

Sits with Sellbrite and Kyozou in the professional tool lane. The brief's finding holds exactly: the survivors in tools are bootstrapped. List Perfectly $0, Crosslist $0, PrimeLister $0, Listing Mirror $0, while Flyp raised $15.8M and now gives the product away. The best exit the lane produced was Sellbrite into GoDaddy at an undisclosed price on $2.9M raised. Nothing here threatens a consumer scan-and-list app and nothing here is threatened by one. Threat to Moe is 1.

Base decomposition

Market and competition 3.0: real category, but the seat is behind ChannelAdvisor, Linnworks, Sellercloud and Zentail, and the lane's proven exit ceiling is a small acqui-hire. Product and wedge 4.5: broad official-API integration set including eBay silver solution provider and Walmart top-tier partner status, plus warehouse and FBA depth. The wedge is service, and service does not compound. Traction 4.5: ten years alive, 8 staff, 3.5-year tenure, $175M monthly GMV claimed. Real but self-reported and small. Unit economics 5.0: bootstrapped to profitability with a sticky product, but human-heavy support caps margin and growth. Team 5.0: genuinely technical CEO with a decade of engineering before founding. One of two founders exited May 2025. Founder trust and disclosure 5.5: open about being a family business, publishes its own metrics, no evasions. Minor ding for a Pricing nav item leading to a feature page.

Mean of 3.0, 4.5, 4.5, 5.0, 5.0, 5.5 is 27.5 over 6 equals 4.58. Adjusted down 0.25 because the business has no mechanism to reach venture scale and one founder has stepped back. BASE 4.33.

The formula line

4.33 + 1.00 (EV, $0 raised) + 1.00 (Geo, USA) + 0.00 (Chain) = 6.33

For Daxos, price and verdict

PASS. Daxos fit 1.5 out of 10. There is no round. A couple who bootstrapped ten years and never took a dollar is not selling primary equity at a seed price. Even if they did, $100k to $500k into an eight-person services-heavy SaaS with a category ceiling of a small acqui-hire cannot return a fund. Its honest use is as evidence for the brief, showing the durable businesses in the tool lane are the ones that never raised.

What could not be checked

Revenue, ARR and margin are undisclosed and no pricing is published, so the $175M GMV claim cannot be converted into a take. The orders and tenure figures are self-reported with no corroboration. Whether Erin Wawok's May 2025 departure was a step back or a split of duties is not visible. Customer count and concentration are unknown. Google Play was not checked because the site surfaces no mobile product.

back to the ranked table

A former Instagram machine learning engineer building SKU-level price resolution across luxury handbag resale, which is the right vertical and the right technical wedge, built on top of inventory that belongs to somebody else.

What it is

A buy-side aggregator for secondhand luxury handbags. Search in plain language ("birkin 25", "Chanel under $3k") and it returns matched listings from more than thirty resellers, benchmarked against historical sales so the shopper sees what the bag is worth and where the cheapest honest copy is (secondsense.co, read 2026-09-16; Harmonic description, 2026-09-16).

The technical claim is item identity resolution: matching the same physical bag across retail, resale and auction sites by colour, size, hardware and condition, and normalising condition grades between sellers (BusinessWire release via AI Journal, 2026-02-24). Named sources and partners include The RealReal, eBay, Rebag, What Goes Around Comes Around and Collector Square.

Not a seller tool, not a cross-lister. A price and search layer on the buyer's side of one high-value vertical.

Founders and team

Chris Lucas, sole founder and CEO from 2025-06. Stanford BS and MS in Computer Science, then Harvard Business School MBA. Senior Machine Learning Engineer at Instagram 2019-08 to 2023-05, Engineer in Residence at Meta in the second half of 2022, two ML engineering internships at Uber in 2017 and 2018 (LinkedIn via Apify harvestapi, 2026-09-16).

Four years as a senior ML engineer at Instagram is exactly the skill this needs: cross-source matching on photographs and messy seller text is a retrieval and embeddings problem. He is technical and the only founder.

Against that: no prior company, no exit, no failure, and the idea was prototyped at business school and went viral before it was a business (Engine StartupsEverywhere profile; PR Newswire, 2025-09-30). Harmonic records headcount 4 and zero person relationships, so no CTO is identifiable; on current evidence the founder is the lead engineer. Location is inconsistent, Harmonic says San Francisco and his LinkedIn says New York, which does not change the geography bonus.

Checks run: Harmonic company record, which returned no person records; the site, behind CloudFront and refused a direct fetch but rendered through WebFetch; the PR Newswire release of 2025-09-30; the BusinessWire release of 2026-02-24 via AI Journal; Crunchbase free profile; Product Hunt; a LinkedIn pull. No adverse record found in the checks run: no litigation, no dissolved entities, no disputed claims, and the Instagram, Meta, Stanford and HBS credentials are consistent everywhere they appear.

Product and traction

Web only. No iOS app was found in the App Store search run 2026-09-16, so the rating-count proxy this market is measured by does not exist here.

Visible: Instagram 9,840 followers, LinkedIn 627 (Harmonic, 2026-09-16), headcount 4. Launched summer 2025; the round closed 2025-09-30 after a three-week raise described as oversubscribed (PR Newswire, 2025-09-30). The creator Alix Earle shared it with a TikTok audience above seven million in April, before launch, which is where the early attention came from.

The site's headline numbers are 1.7M or more handbags indexed, $3.3Bn or more analysed, $341M savings found (secondsense.co, 2026-09-16). Computed marketing figures, not users, revenue or GMV, none of which are disclosed anywhere.

Business model and unit economics

Not disclosed in either press release or on the site, which is itself the finding. The shape, comparison then a click out to The RealReal, eBay, Rebag or WGACA, points to affiliate revenue.

If it is affiliate, the economics are the best in this set by a distance. Handbag order values run from roughly two thousand to well over ten thousand dollars, so a single-digit rate is a payout in the hundreds per conversion against near-zero marginal search cost. Nothing else in this market gets paid that much per event.

Every dollar is granted by the partners. The inventory and the feed are theirs, so the moment a large reseller decides price comparison hurts its margin, the feed stops. The same platform dependence cross-listers face, transposed to the buy side.

Position in this market

The tool lane has a hard ceiling, and where this thesis compounded it was vertical with liquid comps. Luxury handbags are the most liquid comp set in fashion resale, the order values fund real infrastructure, and identity resolution is hard enough to be worth solving. That is the right half of the map.

The wrong half is the position on it. Secondsense owns no supply, no demand and no transaction, only a matching graph over other people's listings. The companies that compounded owned the data and the transaction. The nearest threat is not another startup but generalist shopping agents converging on this comparison for free, plus the resellers' own published price indices.

Base decomposition

• Market and competition 4.5: a real vertical with liquid comps and high order values, offset by an aggregator position over partner-controlled feeds and free AI comparison arriving from the platform layer.
• Product and wedge 5.0: identity resolution across colour, size, hardware and condition with grades normalised between sellers is genuinely hard and correct for this vertical.
• Traction 3.5: about fourteen months live, no app and so no comparable rating count, 9,840 Instagram followers, one creator moment, headline stats counting listings rather than users or revenue.
• Unit economics 4.5: high-AOV affiliate is the strongest revenue shape in this set, but undisclosed, unproven and at the partners' discretion.
• Team 6.0: Stanford CS BS and MS, four years senior ML at Instagram, HBS, exactly matched to the problem. Solo, first company, team of four.
• Founder trust and disclosure 4.0: credentials verify everywhere. Deductions for unsupported "patented AI" language with no patent number located, "the only platform" framing that is overstated, and monetisation disclosed nowhere.

Mean = (4.5 + 5.0 + 3.5 + 4.5 + 6.0 + 4.0) / 6 = 27.5 / 6 = 4.58. No adjustment: the credit for a vertical with liquid comps and the discount for depending on other people's inventory already sit inside the market and unit economics parts. Base = 4.58.

Formula

4.58 + 0.75 (EV, Harmonic funding_total $2.0M, $1M to $3M band) + 1.00 (Geo, USA) + 0.00 (Chain) = 6.33.

For Daxos, price and verdict

Harmonic records no disclosed valuation and gives an ESTIMATED $10M post, derived mechanically from a $2M seed at typical 20% dilution and marked ESTIMATED. Treat it as a floor indication only. That round closed 2025-09-30, roughly a year ago, so the live question is a seed extension or a Series A at a step-up, not entry at $10M.

At $10M or above a $100k to $500k check buys one to five per cent. Returning a fund requires Secondsense to become the price layer for all luxury resale and to monetise the transaction rather than the referral, and requires resellers to keep feeding a product whose purpose is to tell their buyers where the same bag is cheaper. That tension is real and no disclosure addresses it.

For engagement: the founder is the second strongest in this set, the vertical has the characteristics that matter, and the matching graph improves with time rather than commoditising. DAXOS = 5.5. VERDICT MEET. The conversation is about monetisation, partner contract terms and whether the identity graph is owned or rented. No check on current disclosure.

Threat to Moe: 1. Buy side, luxury only, web only, no seller tooling.

What could not be checked

Revenue, monetisation, affiliate rates, users, sessions, conversions, GMV influenced. Whether any patent exists; the claim appears only in the company's own releases and no number was located. Whether the reseller relationships are contracted or scraped. Whether it is raising and at what price: currently_raising is null. Headquarters, which sources disagree on. Any team member other than the founder.

back to the ranked table

Ten years bootstrapped to ten staff at the highest prices in the category, with a real community moat and no native AI capability, and it has never wanted outside money.

What it is

A cross-posting and listing management tool for people who sell online for a living. Lists and crossposts to Poshmark, Mercari, Instagram, Shopify, eBay and more (Harmonic external_description, pulled 2026-09-16). Browser-extension and web first, no native mobile app. Harmonic classifies it Business (B2B), correctly. The feature matrix on the pricing page is the deepest in this batch: multi-account support, Poshmark sharing and delisting tools, bulk import and bulk edit, inventory labels, stale-listing detection, auto delist, sub accounts, Shopify integration, custom titles and footers per marketplace, CSV export and sales analytics.

Founders and team

Two co-founders, both career resellers who built the tool for their own business first. That origin is the company's real asset.

Amanda Morse, co-founder since 2016-08 and CEO since 2019-03. University of Michigan BA. Co-founder and, in her own words, "Tech Developer" at The Clothing Vault, an online clothing retailer, since 2004, so more than twenty years selling online. Marine Corps reservist 1993 to 2001 (Harmonic person 159056783).

Clara Albornoz, co-founder, Co-CEO from 2020-11 to 2023-11, then Chief Strategy Officer, and since 2025-04 Chief Strategy Officer for Marketing and Community. Founder and President of The Clothing Vault 2011 to 2019, described as a multi-channel retailer selling across its own site, eBay, Etsy and Facebook. In 2016 she founded List Clothing, which her own record calls "a custom solution for The Clothing Vault and my first MVP that still continues to power it," which is List Perfectly's origin. Before reselling she spent fifteen years in mortgage and retail banking at Countrywide, Chase, Bank of America and Citi, and trained as a civil law notary in Argentina (Harmonic person 16753104).

On the technical question I am marking unknown rather than yes or no. Morse's only technical credential is a self-assigned "Tech Developer" title at her own small retailer. Neither founder has an engineering degree, an engineering employment history, or a GitHub presence that surfaced in the checks run, and no CTO or lead engineer appears in Harmonic's people list for the company. With ten staff there is engineering somewhere, but it is not identified.

Checks run: Harmonic company record and both founder person records, company pricing page, US App Store search, one general web search covering litigation, trademark disputes and platform conflicts. No litigation, dissolved entity, disputed claim or inflated title surfaced. No adverse record found in the checks run. Albornoz's title has shifted twice since 2023, from Co-CEO to CSO to CSO for Marketing and Community, which reads as a narrowing scope rather than anything adverse.

Product and traction

No App Store presence. A search for "list perfectly" on the US store on 2026-09-16 returned no company app, and the pricing page carries no Apple or Google Play links. Mobile is browser-based, marketed as "Mobile Access" and included on every tier. That removes the one traction proxy that makes this market comparable, so List Perfectly cannot be measured against SellRaze's 39,656 ratings or Hero's 2,060 on the same axis.

What can be verified: ten years of continuous operation, ten headcount, and zero outside funding across zero rounds (Harmonic, 2026-09-16). A ten-person payroll funded entirely by subscriptions at $29 to $99 a month implies a business doing low seven figures. That is a real company. It is also, after a decade, smaller than Vendoo, which reached 39 staff in the same lane on $125,000.

The distinctive product is Listing Party, a community and live-session product bundled into every tier. Competitors can copy a crosslisting feature in a sprint. They cannot copy a decade of relationships with the reseller community, and this is the one durable moat in the batch that is not code.

Business model and unit economics

Four tiers: Simple $29, Business $49, Pro $69, Pro Plus $99 and up, all monthly (listperfectly.com/pricing, 2026-09-16). Crosslisting and listings are unlimited on every tier. What is metered is the expensive part: background removals at 25 a month on Simple and 50 on Business, and AI-generated listings and barcode scans at the same caps.

These are the highest prices in the category, roughly double Vendoo's entry tier and eight times SellRaze's. The economics are the best in the batch for the same reason: a seller paying $69 a month is running a business, treats it as cost of goods, and does not churn on a closet clean-out. Ten years of self-funding is the proof.

One telling detail in the feature list: a "Google Lens Pricing Tool" and an "eBay Pricing Lookup Tool." List Perfectly does not run its own vision model for valuation. It routes users to Google Lens. That is capital-efficient and it is also the whole competitive problem, because the thing SellRaze, Hero, Haz and Moe are building is exactly the model List Perfectly outsources.

Position in this market

List Perfectly is the archetype the category brief names: a bootstrapped cross-lister that outlasted the funded ones. Flyp raised $15.8M, has not priced a round since 2022-05, and now gives the product away. List Perfectly raised nothing and charges $29 to $99. That comparison is the single clearest statement of what this lane is.

Its exposure is the same as everyone's. Poshmark suspended cross-lister users twice in 2025 and List Perfectly's Poshmark sharing and delisting tools sit directly in that blast radius. Mercari's 2024-08 change broke cross-listers without notice. The difference is that a ten-person bootstrapped company with no burn can absorb a platform shock that would kill a venture-funded competitor mid-runway.

The strategic risk is slower and more serious. The wedge in this market has moved to vision AI, and List Perfectly is a decade-old listing manager wrapping Google Lens. Its customers are the ones least likely to switch, which buys time, but the direction of travel is against it.

Base decomposition

Market and competition 3.0: the hard category ceiling, in its most crowded lane, against a free well-funded competitor. Product and wedge 4.5: deepest feature matrix in the batch and a genuine community moat in Listing Party, offset by no native AI valuation capability. Traction 4.5: ten years and ten staff on zero funding is real, but smaller than Vendoo on the same clock and with no verifiable public traction proxy. Unit economics 6.0: highest prices in the category on a professional buyer, unlimited listings with the costly AI features metered, proven by a decade of self-funding. Team 3.5: outstanding domain authenticity, no identified engineering leadership, and a co-founder whose scope has narrowed twice since 2023. Founder trust and disclosure 5.5: fully public pricing, a public founder story, ten years of continuous operation, nothing adverse.

Mean = (3.0 + 4.5 + 4.5 + 6.0 + 3.5 + 5.5) / 6 = 27.0 / 6 = 4.50. Adjust -0.25 because the category's wedge has moved to vision models and this company has neither the capability nor an identified technical leader to build one. Base = 4.25.

The formula line

COMPANY = 4.25 Base + 1.00 EV + 1.00 Geo + 0.00 Chain = 6.25

Note that the +1.00 EV bonus here rewards a company for never having raised, which is accurate as capital efficiency but should not be read as headroom. List Perfectly has not raised because it does not want to, not because it is early.

For Daxos, price and verdict

There is nothing to buy. Ten years, zero rounds, zero recorded investors, and Harmonic classifies the stage as VENTURE_UNKNOWN because no financing event has ever been recorded. Two founders own a profitable business that pays them, in a category whose best realistic outcome is a mid-eight-figure tool acquisition. Selling a seed slice at a price that could return any part of a fund would require them to accept a valuation far below what the cash flows are worth to them, and they have declined that trade for a decade.

Even if they took a cheque, the maths fails on the exit side. A $100k to $500k position in a company whose ceiling is a modest strategic acquisition, split two ways after ten years of founder ownership, returns a multiple that does not move a fund. Daxos brings nothing this team needs. They have the community, the customers and the cash flow.

VERDICT: PASS. Keep the card as a benchmark. When a funded company in this batch claims the tool lane supports a venture outcome, List Perfectly is the counterexample: the most price-disciplined, longest-surviving operator in the lane chose never to raise at all.

What could not be checked

Revenue, subscriber count, churn and profitability: no source, all inferred from headcount and pricing. Who leads engineering: no CTO or engineer appears in Harmonic's people list. Whether any mobile app exists on Google Play. Listing Party participation numbers. Which marketplace integrations use official APIs versus browser automation. Whether the founders have ever received or entertained an acquisition approach.

back to the ranked table

A B2B inventory and pricing tool for estate liquidators, with by far the best team in this batch and by far the weakest traction, sitting upstream of every item before it reaches a marketplace.

What it is

Software for estate sale companies, liquidators and antique dealers. A team walks a house, photographs items, and the app identifies each one, writes a description, researches sold comparables, suggests a price, and cleans up the photograph. The photo enhancement is specific to the job: changing backgrounds, removing clutter, removing shadows, removing mirror reflections, folding blankets, putting clothes on mannequins. It then produces listings, flysheets and reports, supports multiple team members on one project, and captures offline for later processing. Sources: https://valuable.app and https://valuable.app/pricing, read 2026-09-16.

The customer is the professional, not the consumer. That is the single most important fact about this company and it separates it from everything else in this batch.

Founders and team

Five co-founders, named on the record by GrepBeat on 2025-04-17 (https://grepbeat.com/2025/04/17/raleigh-based-valuable-assesses-antiques-and-estate-items-straight-off-the-phone/): Bill Goodwin, Amy Wolpert, John Boezeman, Magdalyn Duffie and John Wolpert.

John Boezeman, CTO, is a former IBM Distinguished Engineer and Master Inventor. Those are internally awarded IBM titles with a small population, and they mean a career of shipped systems and issued patents. He is a real technical founder.

John Wolpert, Chief Product Officer, is a former IBM product executive across Watson, Blockchain and Web, the founder of the mobile ridesharing company Flywheel, and the author of The Two But Rule. Long product career, one venture-backed consumer company behind him.

Bill Goodwin, CEO, is an antiques collector and dealer and the owner of Goodwin Executive Search and Wm Carroll. GrepBeat reports he funded and principally founded the project. He brings southeastern market access and capital, not software experience, and he is still running a search firm.

Amy Wolpert, Chief Customer Officer, is the estate sale operator whose workday produced the idea: her husband watched her spend hours inventorying a house. She is the customer, in the building.

Magdalyn Duffie, Chief Creative Officer, is a designer. Patrick McGowan, UX lead, did fifteen years of UX and product management at IBM. Noah Yarborough is a full stack developer on support. Katie Schulist runs training.

The discount: eight people carry titles on https://valuable.app/meet-the-team but Harmonic puts headcount at 5, so several are part time, and the CEO has another business. A C-suite that is mostly ex-IBM executives is strong on judgment and thin on the number of hands actually shipping.

Checks run: Harmonic company enrich on valuable.app (no person records attached), the company site including team, pricing and press pages, the App Store and Google Play listings, GrepBeat's April 2025 profile, and web search on the founder names. No adverse record found in the checks run: no litigation, dissolved entity or disputed claim surfaced. The Apify LinkedIn actor was not run, because the team page and the press profile already gave checkable histories for every named person.

Product and traction

Traction is the problem. 28 iOS ratings in twenty-two months, and about nine Google Play reviews at 4.3. Harmonic shows web traffic of 5,176 and an Instagram following of 5,598 against 112 on LinkedIn. Valuable presented at Venture Atlanta 2025 and was covered by GrepBeat and myLifeSite (https://valuable.app/press, read 2026-09-16).

Ratings undercount B2B seats, since a crew of five sharing one project produces one rating at most, so the real customer number is probably higher than 28. But it is not much higher, and twenty-two months is long enough that it should be. The headline claims, "USED BY THE NO. 1 ESTATE SALE COMPANY IN THE U.S." and "the largest, most profitable estate sales companies", name no customer.

No marketplace listing integration is named anywhere on the site. Valuable prepares listings; it does not post them. That keeps it clear of the cross-lister ban risk that has hit this category, and it also keeps it out of the transaction.

Business model and unit economics

The pricing model changed. In April 2025 GrepBeat described tiered subscriptions at $4.99 a month for 50 items, $9.99 for 200 and $19.99 for 5,000. As of 2026-09-16 the site sells credit packs instead, with everything else free forever: Home Bundle $19.99 for 400 credits, Estate Bundle $99.99 for 2,000, Pro Bundle $399.99 for 8,000, credits never expire. An AI-created item costs 4 credits and photo enhancement costs 4, so every AI action is a flat $0.05 and a fully processed item is about $0.10 to $0.20.

Consumption pricing against per-item inference cost is the right shape and almost certainly positive gross margin. Two caveats. Moving off subscriptions after a year usually means the subscriptions were churning. And because Valuable takes no share of the goods sold, revenue is capped at usage in a business that is seasonal and lumpy: a liquidator who runs eight sales a year buys credits eight times.

Position in this market

The brief's finding is that seller tools have a hard ceiling and only demand networks produce venture outcomes. Valuable sits in the tool lane, but in the professional half of it, where there is no free first-party substitute. eBay's photo-to-listing, Facebook's AI listings and Google Lens all serve one consumer with one item. None of them handles a five-person crew inventorying a four-bedroom house offline in a morning, and none prints a flysheet. That is a real product with a real buyer, and it explains the testimonial on the homepage: "After using Valuable, I'll never go back to using Google Lens again."

The strategically interesting part is position. Valuable touches secondhand inventory before it reaches any marketplace, at the moment of supply creation. In this category the money has been in demand networks and in businesses upstream of the camera. Valuable is upstream. It is not yet doing anything with that.

Base decomposition

• Market and competition: 4. A defensible B2B niche with genuine willingness to pay and no free first-party alternative, but a small one. The US estate liquidation trade is thousands of operators, not millions, and EstateSales.NET, EBTH and newer entrants already serve it.
• Product and wedge: 5. Multi-member on-site capture, offline queueing, job-specific photo enhancement, flysheets and printing. This is workflow depth that a general model wrapper cannot copy in a weekend.
• Traction: 2. 28 iOS ratings and about 9 Play reviews in twenty-two months, with unnamed marquee customers. Whatever the seat count is, it has not compounded.
• Unit economics and business model: 4. Clean consumption pricing at roughly $0.20 a finished item, plausibly gross margin positive, but capped at usage, seasonal, with no participation in the value of goods sold and a pricing pivot that hints at churn.
• Team: 6. An IBM Distinguished Engineer as CTO, an IBM product executive and Flywheel founder as CPO, a working estate liquidator as CCO, and domain access through the CEO. This is the only team in the batch that could plausibly build and sell a real company. Held below 7 because most of them are part time against a headcount of five.
• Founder trust and disclosure: 4. Named, checkable team, consistent founding story, honest footnote defining what "unlimited" means. Against that: unnamed marquee customers, an undisclosed friends-and-family round that does not appear on Harmonic, and a misspelled sales address on the pricing page (sales@valuaple.app).

Mean = (4 + 5 + 2 + 4 + 6 + 4) / 6 = 25 / 6 = 4.17. Adjusted up 0.08 to 4.25 for the one structural advantage nothing else in this batch has: a paying professional buyer with no free substitute, reached before the goods hit a marketplace.

The formula

Base 4.25 + EV 1.00 (Harmonic shows $0 raised; even taking the reported friends-and-family round at face value it falls in the $0 to $1M tier) + Geo 1.00 (Raleigh, North Carolina) + Chain 0.00 = COMPANY 6.25.

For Daxos

Worth a conversation, not yet a check. As of April 2025 the team said it intended to move from bootstrapped to pre-seed, so it is probably reachable and probably raising. At the revealed prices in this category, a pre-seed here would likely be $500k to $1.5M on a $4M to $8M post.

The honest case against writing: even executed perfectly, estate liquidation software is a good small business. A few thousand US operators buying credit packs eight times a year does not return a seed fund, and the category brief is explicit that the tool lane has never produced a venture outcome. Valuable also takes no share of what it prices, so it captures none of the upside in the goods it touches.

The case for a conversation: the team is genuinely good, the CTO is a real engineer, the CCO is a real customer, and the company sits upstream of the whole supply chain, seeing what is in the house before anyone else does. If the answer to "what do you do with that data and that position" is a demand network, a wholesale channel, or a financing product against inventory, that is a fund-returning shape and it is worth an hour to find out. If the answer is "sell more credit packs", pass.

Price discipline: a $100k to $250k check only at or below roughly a $6M post, and only if the plan extends past software fees. Daxos adds little operationally here beyond introductions. DAXOS 5.0. VERDICT MEET.

What could not be checked

The friends-and-family round amount, the customer count, revenue, retention and credit-pack repurchase rate are all private and not on Harmonic. The claim to be used by "the No. 1 estate sale company in the U.S." names no company and could not be verified. Whether the subscription-to-credits change was driven by churn or by customer preference is inferred, not established. Whether the company is currently raising is unknown: Harmonic's currently_raising field is null.

back to the ranked table

Nine years old, 39 staff on $125k raised, the best real revenue in the cross-lister lane and the clearest proof that this lane tops out as a durable small business rather than a venture outcome.

What it is

A cross-listing and inventory tool for online resellers. Bulk upload and manage inventory across many marketplaces "including those without APIs," which is the company's own wording for browser automation (Harmonic external_description, pulled 2026-09-16). Feature set on the site: crosslisting, sale detection and auto-delist, bulk actions, inventory management, multi-quantity, send offers, analytics, mobile app, marketplace sharing, plus free calculators for fees and profit margin. Harmonic classifies it Business (B2B), which is right: this is a tool for people who resell for a living, not for a closet clean-out.

Founders and team

Four co-founders, all four still listed as current after nine years. That is unusual and it is the most interesting fact about this company.

Thomas Rivas, Founder and CEO since 2017-06, University of Maryland. Before Vendoo: freelance Spanish and English translator 2016 to 2019, a four-month software development internship at Recentia Health writing Angular in 2017, bank teller at BB&T, assistant coach at Club Pachuca (Harmonic person 372927). Not an engineer by training.

Chris Amador, Co-Founder, CTO and Head of Product since 2017-06, George Mason. Front-end developer at Gallup 2016 to 2018, owner of Hotfix, an identity and development studio, 2018 to 2020, web and graphic design roles before that (Harmonic person 372928). Technical, but design and front-end rather than systems or data.

Benjamin Martinez, Co-Founder since 2017-01, University of Maryland. Harmonic shows no experience outside Vendoo.

Josh Dzime-Assison, Co-Founder and CMO since 2017-05, University of Delaware. Has run Get Dizzy, a styling and image consulting store, since 2007 (Harmonic person 160006806). He is the domain seller on the team.

Checks run: Harmonic company record and four person records, company pricing and marketplaces pages, App Store lookup for both apps, one general web search on the CEO and the company covering litigation and platform suspensions. No litigation, dissolved entity, disputed claim or inflated title surfaced. No adverse record found in the checks run.

Product and traction

US App Store, checked 2026-09-16. Main app id 1612168777: 2,784 ratings, 4.53 average, released 2022-03-29, last updated 2026-08-20. A second app, Vendoo Go, id 6746722923, released 2026-01-21 and updated 2026-09-08, has 0 ratings. Vendoo Go is the tell: nine years in, they are launching a separate lighter app, which reads as a move toward the consumer scan-and-list lane that SellRaze, Moe and Haz occupy.

The real traction number is not the app. It is 39 headcount (Harmonic, 2026-09-16) sustained for nine years on $125,000 of outside money. Nothing else in this competitor set can show that. Payroll for 39 people, even heavily distributed, implies several million in annual revenue. Vendoo is the revenue leader in the tool lane and it got there without venture capital, which is exactly the pattern the category brief describes for List Perfectly, Crosslist and PrimeLister.

Business model and unit economics

Subscription, three tiers plus enterprise, 14-day free trial. Starter $14.99 a month or $12.49 billed yearly, unlimited items, all marketplaces, sale detection and auto-delist. Growth $29.99, adds AI listing enhancement, bulk actions to 240 listings, 300 PhotoRoom background removals. Pro $59.99, adds auto-send offers on six-plus marketplaces, marketplace sharing for Poshmark, Depop and Grailed, 1,500 removals, listing videos. Enterprise on contact for businesses creating over 1,000 listings a month (vendoo.co/pricing, 2026-09-16). The site pitches the subscription as tax-deductible, which tells you the buyer files a Schedule C.

These are the best unit economics in the set. A pro reseller paying $29.99 a month against thousands in monthly gross merchandise value has low price sensitivity and low churn, and the enterprise tier pushes into warehouse operators where the contract value is real. Costs are PhotoRoom API calls and, increasingly, model inference, both metered and capped by tier.

Position in this market

Vendoo sits in the crowded cross-lister lane alongside the $0-raised survivors and a Flyp that raised $15.8M, never priced a round after 2022-05, and now gives the product away. Vendoo's answer has been to go upmarket into enterprise rather than compete on free, which is the correct move and still leaves it inside a lane with no venture exit.

Platform dependence is live and documented. An eBay community thread titled "Third-Party Crosslisters Like Vendoo, FLYP can't transfer listings" (community.ebay.com, thread 35034415) shows the breakage is public and user-visible. Against the backdrop of Mercari's 2024-08 change, the two Poshmark cross-lister suspension waves in 2025, and eBay's 2026-02 ban on LLM-driven agentic bots, a company whose own marketing says it supports marketplaces "without APIs" is describing its main risk.

Base decomposition

Market and competition 3.0: hard category ceiling, and this is the most crowded lane in it, with a free well-funded competitor and several profitable bootstrapped ones. Product and wedge 4.5: mature and broad, the enterprise tier is a genuine differentiator, but there is no defensible wedge a competitor could not copy. Traction 5.0: 39 staff on $125k for nine years is strong revenue evidence; 2,784 iOS ratings is modest, though a desktop-first B2B tool understates on mobile. Unit economics 5.5: pro-seller subscription at $14.99 to $59.99 plus enterprise, sticky buyer, metered costs. Best in the set. Team 4.0: four founders intact for nine years, which is real durability, but no engineer with systems depth and no prior exit. Founder trust and disclosure 5.0: transparent pricing, long tenure, nothing adverse found.

Mean = (3.0 + 4.5 + 5.0 + 5.5 + 4.0 + 5.0) / 6 = 27.0 / 6 = 4.50. Adjust -0.25: nine years and a YC batch have already produced the answer on whether this breaks out, and it did not. Base = 4.25.

The formula line

COMPANY = 4.25 Base + 1.00 EV + 1.00 Geo + 0.00 Chain = 6.25

For Daxos, price and verdict

Vendoo is almost certainly not raising. The last event Harmonic records is the YC accelerator cheque dated 2022-01-01, four and a half years ago, and Harmonic flags its own $625k estimated valuation as stale for that reason. A company that has built to 39 people on $125k has no reason to sell primary equity cheaply, and if it ever does raise it will be priced off a revenue multiple in a growth or secondary conversation, not as a seed round. A $100k to $500k cheque has no natural entry point here.

Even at a friendly price the outcome maths fails. The exits in this lane are mid-eight-figure tool acquisitions at best, and Vendoo would be splitting that across four founders and any new holder. That does not return a seed fund. Daxos adds nothing a nine-year-old team with a YC network needs.

VERDICT: PASS. The useful output from this card is not the company, it is the benchmark. Vendoo is the ceiling of the cross-lister lane made concrete, and any consumer tool in this set pitching a venture outcome has to explain why it beats a company that reached 39 staff and real revenue on $125,000.

What could not be checked

Actual revenue, ARR, subscriber count, churn and profitability: no source, all inferred from headcount. Whether the 39 headcount is employees or includes contractors. Google Play ratings and installs. Which marketplace integrations run on official APIs versus browser automation, beyond the company's own "including those without APIs" phrasing. Enterprise contract values and customer count. Whether Vendoo Go is a strategic consumer push or an experiment.

back to the ranked table

Eighteen-year bootstrapped two-founder LLC running twelve collection-cataloging apps with 35,611 iOS ratings and a one-time-purchase model, the clearest proof that this lane supports a lifestyle business and nothing larger.

What it is iCollect Everything publishes collection cataloging apps for iPhone, iPad, Android, Mac and Windows. Scan a barcode or search, add the item, track value, sync across devices. The flagship covers more than 25 collectible types, from movies and comics to Hot Wheels, LEGO, wine, coins, Funko, watches, stamps and trading cards, plus custom collections (App Store description, 2026-09-16). It catalogs and values. It does not list, does not cross-post, and takes no cut of a sale.

A Harmonic record exists but is near empty: funding_total 0.0, zero rounds, null status NONE_ANNOUNCED, no location, founding date, headcount or person records (Harmonic enrich, 2026-09-16).

Founders and team Formed in summer 2008 by Ethan C. Allen and Chris Weier, one month after the App Store opened. Incorporated as Hooked In Motion, LLC, renamed Sort It! Apps, LLC, then iCollect Everything, LLC after a full rewrite to support custom collections (icollecteverything.com/about, 2026-09-16).

Ethan C. Allen, CEO and co-founder, lives in Windermere, Florida, claims 30 years in tech and has written software since 1995. He is the named developer on the apps and signs the pricing and release posts. Technical, and the operator. Chris Weier, co-founder, lives in Fife, Washington, with close to 20 years in software. Two technical founders, eighteen years on one problem, no outside capital, no disclosed employees. There is no separate CTO because there is no separate anyone.

Checks run: Harmonic company record (no person records returned), the about page and pricing post, App Store publisher listings, web search on the legal entity name. No adverse record found in the checks run. LinkedIn scraping was not run.

Product and traction US App Store, pulled 2026-09-16. Everything: Inventory 13,673 at 4.64 (rel 2018-06-24), Movies 8,045 at 4.59, Books 4,111 at 4.58, Comic Books Keys 3,084 at 4.56, Toy Cars 2,140 at 4.49, Video Games 1,843 at 4.61, Music 1,170 at 4.55, Action Figures 700 at 4.48, Funko 527 at 4.63, Trading Cards Scanner 280 at 3.90, Stamps 27 at 3.59, Ornaments 11 at 4.27. Total 35,611. Oldest apps date to 2011-11-03.

Every app shares a last-update date of 2026-06-02, one release across a shared codebase and nothing since. And the trading card app has 280 ratings against Collectr's 64,990 on a free app in the same vertical: the one category here with liquid comps and real resale money is the one where the generalist is invisible.

The App Store listing claims official use by over 2,000 Tennessee teachers across five school districts, and the about page names eBay, Swap.com, USPS and Insure as collaborators (2026-09-16). The school use is the more interesting revenue fact and has nothing to do with resale.

Business model and unit economics Mostly one-time purchase. Lifetime Unlimited Items unlocks at $29 for one collectible type or $59 for all, prompted after 30 items. Desktop needs a separate Pro subscription at $79 per year (icollecteverything.com pricing post, 2024-10-26, accessed 2026-09-16). That caps lifetime value hard and gives the company no reason to care whether a user ever sells anything. No take rate, no transaction. The 30-item free tier is the funnel.

Position in this market The bootstrapped-survivor pattern the brief describes, moved from cross-listing into cataloging. It sits with List Perfectly, Crosslist and PrimeLister at zero raised, and has outlasted the funded graveyard: Gone $3.8M, Lovd $4.1M, Gemr $7.4M. Eighteen years, no capital, 35,611 ratings, still shipping. A genuinely good outcome for two people and a terrible one for a fund. No threat to a scan-and-list product: it builds inventory lists, not listings.

Base decomposition Market and competition 3.0: "what is this worth" is free at the OS layer via Google Lens and Apple Visual Intelligence, and vertical apps win every category that carries money. Product and wedge 5.0: 25-plus collectible types under one roof is a real differentiator, but breadth is exactly what loses to depth here. Traction 6.0: 35,611 ratings, second-largest in this comparison set, accumulated over fifteen App Store years. Unit economics 3.0: one-time $29 to $59, no take rate, unknown Pro attach, no way to grow revenue as a collection grows. Team 4.0: two capable technical founders shipping for eighteen years, no evidence of anyone else, no venture ambition on display. Founder trust and disclosure 6.0: plain public history including two prior entity names, pricing published openly, no adverse findings. Mean of 3.0, 5.0, 6.0, 3.0, 4.0, 6.0 is 27.0 divided by 6 equals 4.50. Adjusted down 0.25 for flat growth, one release so far in 2026, and the 280-rating trading card app proving the generalist cannot hold a liquid vertical. Base 4.25.

The formula 4.25 + 1.00 (EV, $0 raised per Harmonic) + 1.00 (USA) + 0.00 (chain) = 6.25. The two-point bonus stack does most of the work; the business is a shade above an adequate median.

For Daxos, price and verdict DAXOS 1.0. An eighteen-year-old two-person LLC with no outside capital is not raising, has no reason to, and offers no structure a $100k to $500k preferred check fits into. Even at the SellRaze benchmark of $500k on a roughly $2.0M Harmonic-estimated post, there is no fund-returning path from a one-time-purchase cataloging app in a category whose only venture outcomes were demand networks. PASS.

What could not be checked Revenue, paid conversion off the free tier, Pro attach rate, Android and desktop install base, headcount beyond the two founders, and the substance of the claimed eBay and USPS collaborations.

back to the ranked table

French B2B recommerce engine pushing retailer dead stock onto Leboncoin, Vinted and eBay, run by a repeat founder pair with a prior exit, but enterprise-sold in France with a trade-sale ceiling.

What it is

Not a consumer cross-lister despite the tag. Speeral sells to brands and retailers: "la plateforme IA qui couvre tout le cycle de revente C2C" (speeral.io, fetched 2026-09-16). Take a retailer's dormant stock, returns and trade-ins, generate listings, price dynamically, handle buyer negotiation, push to Leboncoin, Vinted and eBay. The retailer hook is that 47 percent of C2C sales generate an in-store pickup, so resale drives footfall rather than cannibalising it, and margin runs 20 to 60 percent better than selling to a liquidator (site figures, unaudited).

Founders and team

Julien Chevalier, CEO, start 2024-06. Harmonic: Founder and CEO of Teester 2016 to 2021, recorded as acquired by Providence Equity; SVP and co-founder of Skeepers 2021 to 2022; Director of Cdiscount Ads Marketplace 2014 to 2016 and Senior Manager Strategic Alliances at Cdiscount 2013 to 2015; Advertising Director at relevanC (Casino group) 2011 to 2013. Business education (ENC Nantes, IDRAC), not technical.

Matthieu Charron, CTO, start 2024-06. Harmonic: CTO and partner at Teester 2016 to 2021, the same company and years, so this is a repeat pair; Engineering Manager at Skeepers 2021 to 2022; co-founder and R&D engineer at Carlili 2015 to 2016; founder and CEO of Troovon 2011 to 2016; engineer at Vrbo and SQLI before that. Master from ESIEE in computing. Technical founder confirmed: fifteen years of engineering and engineering leadership. Troovon ran five years with no visible exit; Matrice Carree 2023 to 2025 closed.

One prior exit together (Teester into the Providence-backed Skeepers rollup). The angel list matters more than the fund: Philippe Brochard, former CEO of Auchan France; Emmanuel Grenier, Cdiscount and Carrefour Group; Guillaume Cavaroc, Meta; Antony Rode (frenchweb.fr 2025-02 and aquiti.fr, read 2026-09-16). That is direct distribution into exactly the buyer they need.

Checks run: Harmonic company and person records, company site, French press (frenchweb, Le Journal des Entreprises, jaimelesstartups), the Aquiti portfolio page, Pappers registry listing for SPIRAL (SIREN 930872064). No adverse record found in the checks run.

Product and traction

Named customers on the site: Boulanger, BUT, Conforama, Monoprix, E.Leclerc, Happy Cash (speeral.io, 2026-09-16). Whether these are signed contracts or pilots is not stated. For French retail those are top-tier logos and the strongest traction signal in this batch.

"Speeral Pro" on the FR App Store (id 6775519707, seller SPIRAL (FR), sellerUrl speeral.io/destockage), free, released 2026-06-21, updated 2026-09-03, 0 ratings as of 2026-09-16. It is a store-staff tool for managing the consignment deposit, not a consumer app, so 0 ratings is expected rather than damning. Harmonic web_traffic 1,189, LinkedIn 1,295 followers. Headcount 12, at least four engineers.

Integrations run through professional-seller channels, not consumer accounts. Vinted Pro and Leboncoin professional accounts are sanctioned routes the marketplaces actively want supply through. That makes Speeral structurally safer than the consumer cross-lister lane, where Poshmark suspends users and Mercari silently breaks integrations.

Business model and unit economics

Pricing is not published. The model is almost certainly a platform fee plus a share of recovered GMV, sold per retail chain. The customer-side gross margin argument is strong (20 to 60 percent better than a liquidator, plus footfall). The company risk is the sales motion: enterprise French retail with long cycles and procurement, funded by a single EUR 1.7M round closed nineteen months ago with 12 people to pay. Either revenue carries part of payroll or runway is thin.

Position in this market

This sits outside the consumer scan-value-list loop. Free photo-to-listing inside eBay, Vinted and Depop does not touch Speeral, because a retailer with 40,000 dormant SKUs will not photograph them one at a time on a phone. That is the right way out of the tool lane: move upstream of the consumer camera to an owner of bulk supply with a budget. The European recommerce enablement field is nonetheless contested by Faume, Reflaunt, Recommerce and Trove, several better funded.

Base decomposition

• Market and competition 5.0: a real and regulation-supported budget line in EU retail, with four or more funded competitors already in it.
• Product and wedge 5.5: specific wedge with the in-store pickup argument, and professional marketplace channels rather than scraped consumer accounts.
• Traction 5.5: six named tier-one French retail logos is far above this batch, discounted because pilot versus contract is unstated.
• Unit economics 4.5: plausible but unpublished; enterprise cycles against one small round and 12 heads.
• Team 6.5: repeat pair, prior exit together, a genuine long-career CTO, angels who run the buyer side.
• Founder trust and disclosure 5.5: customers and metrics named openly, no site claim conflicting with Harmonic.

Mean = (5.0 + 5.5 + 5.5 + 4.5 + 6.5 + 5.5) / 6 = 32.5 / 6 = 5.42. Adjust -0.25 for a crowded funded European recommerce-enablement field where the realistic outcome is a modest trade sale rather than a network. Base = 5.17.

Formula

5.17 + 0.75 (EV, $1.76M in the $1M to $3M band) + 0.25 (Geo, EU) + 0.00 (Chain) = 6.17.

For Daxos, price and verdict

Harmonic shows currently_raising null and one round, closed 2025-02-01. The valuation_info of $9M is ESTIMATED from round size at assumed 20 percent dilution, not reported. A French SAS in Nantes, seeded by a regional fund and French retail angels, is a hard place for a US seed fund to add value: Daxos has no distribution into Boulanger or E.Leclerc, no legal or tax convenience in a euro-denominated SAS, and no information advantage. The outcome shape is a trade sale to a recommerce group or retail software buyer, which does not return a $100k to $500k check at fund scale. DAXOS = 3.5. VERDICT PASS.

This is the best team in the batch and the only company here selling to a buyer with a budget. If it were US-based and raising, the answer would differ.

What could not be checked

Revenue, contract values, and whether the six named retailers are paying or piloting. Pricing is not published. A CFNEWS headline dated after the pre-seed suggests further investors may have come in, but Harmonic records one round at $1,761,565, so no later round is counted and the discrepancy is unresolved. LinkedIn was not scraped. Google Play presence was not visible from the site.

back to the ranked table

The only company in this batch with genuine consumer pull and a top-decile team, but it is a buy-side affiliate play with unproven economics, a settled retailer lawsuit and a Solana token.

What it is

Dupe.com is a visual search engine that takes an expensive furniture, fashion or beauty item and finds cheaper visually similar alternatives across retailers, as a website, a browser extension and an iOS app. The legal entity is Carrot Cart, Inc. (App Store seller name, 2026-09-16); the company operated as Carrot before the rebrand and Harmonic still lists the LinkedIn handle as addtocarrot. The site claims "Trusted by 20+ million shoppers" (dupe.com, 2026-09-16). Revenue is affiliate commission.

This is buy-side discovery. No listing, valuation or cross-posting function, so it does not touch the seller-tool lane.

Founders and team

Bobby Ghoshal, Co-Founder and CEO. Co-founder of Candid, the direct-to-consumer clear-aligner company, before that Head of Digital Design and Growth at WeWork, and earlier the founder of Flud. Design and growth rather than engineering. Candid cuts both ways: he helped build a well-funded consumer brand, and it contracted hard after its studio model failed.

Ramin Bozorgzadeh, Co-Founder and CTO. BA in Computer Science, SUNY Stony Brook, 1998 to 2002. Head of Engineering at WeWork, prior time at Candid, and a public profile describing growing engineering teams from 10 to 600-plus. A genuine technical co-founder with scale experience, rare in this set.

Adverse record found, and it is material. Williams-Sonoma, Inc. v. Carrot Cart, Inc. d/b/a Dupe.com, S.D.N.Y. 1:24-cv-06597, filed July 2024, alleging copyright infringement, false advertising and unfair competition over product photography and disparaging social ads (dockets.justia.com; thefashionlaw.com). Dupe.com settled on confidential terms (Bloomberg Law, news.bloomberglaw.com/ip-law/dupe-com-settles-williams-sonomas-copyright-false-ads-lawsuit). Separately, the site footer links "$DUPE Dupe on Solana", a token (dupe.com, 2026-09-16).

Checks run: Harmonic company record, web search on both founders, federal docket search, App Store seller entity, company site and footer. Nothing further surfaced.

Product and traction

US App Store, 2026-09-16: 5,828 ratings at 4.82 average, released 2025-11-13, last updated 2026-08-25. Roughly 5,800 ratings in ten months is the strongest ratings velocity in this set by a wide margin. Harmonic web traffic 1,176,000 monthly. Instagram 407,941, Twitter 8,430, LinkedIn 1,891. Headcount 24. The consumer demand here is observable, not asserted.

Business model and unit economics

Affiliate commission. The structural problem is that the product's promise is spending less, so it monetises the cheap end of the basket, and furniture affiliate rates are thin. No revenue, GMV or take rate is disclosed. Twenty-four people on $5.5M cumulative is a burn that commission on discount furniture will struggle to carry, and the Solana token reads as a search for a second monetisation surface, which is itself a signal about the first.

Position in this market

Dupe sits on the demand side, where this category's venture outcomes have been. But those outcomes owned liquidity. Dupe brokers clicks into retailers who, as Williams-Sonoma showed, would prefer it did not exist, while Google Lens does visually similar search free at the OS layer. Its real asset is the cultural position of the word "dupe" and a 408k-follower audience. That is a brand, not a moat.

Base decomposition

• Market and competition: 4. Huge consumer shopping market, but the function is free at the OS layer and the supplier side is hostile.
• Product and wedge: 5.5. Working visual search across three verticals, web plus extension plus app, and an audience that arrived on its own.
• Traction: 6. 5,828 ratings in ten months at 4.82, 1.18M monthly visits, 408k Instagram. Best in the batch by an order of magnitude.
• Unit economics: 3. Affiliate revenue on deliberately cheap purchases, 24 heads, nothing disclosed, and a token hinting the core model is not carrying the company.
• Team: 7. Repeat consumer founders, WeWork and Candid pedigree, a CS-trained CTO who has scaled engineering, credible cap table.
• Founder trust and disclosure: 3.5. A settled federal copyright and false-advertising suit, an unverifiable 20M shopper claim, and a Solana token.

Mean = (4 + 5.5 + 6 + 3 + 7 + 3.5) / 6 = 29 / 6 = 4.83. No adjustment: the top-decile team and the best traction in the set are offset almost exactly by the litigation history and the token. Base 4.83.

Formula

4.83 Base + 0.25 EV ($5.5M, the $5M-$10M tier) + 1.00 Geo (USA) + 0.00 Chain (Solana, not Berachain) = COMPANY 6.08.

For Daxos

MEET, no check. Harmonic shows currently_raising null and the last round closed 2026-03-18, so nothing is open, and Harmonic carries no valuation figure. Even with an allocation, $100k to $500k behind M13 and Kindred at a New York consumer seed price buys almost no ownership, in an affiliate referral business whose economics are unproven and whose suppliers have already sued it once.

The meeting is worth taking for information. Dupe is the only company in this comparison set with real consumer pull in the value-shopping adjacency, and Ghoshal is the one founder who can explain why the demand side works and the tool side does not. Fit 4 out of 10.

Threat to Moe: 1. Buy side, no listing product, different verticals.

What could not be checked

Revenue, take rate, referred GMV and monthly actives are undisclosed. The Williams-Sonoma settlement is confidential, so whether it constrains Dupe's advertising or image use is unknown. The $DUPE token's structure, supply and any company treasury position are not visible from the site. Android installs were not visible, and the split of the $5.5M across the two Harmonic-recorded rounds was not disclosed.

back to the ranked table

A bootstrapped sports card scanner with 11,304 ratings built by one computer science student from age eighteen, now fourth in its own vertical behind far better funded rivals and visibly losing on accuracy.

What it is

An iPhone app for cataloguing sports cards. Scan a card, OCR and AI pull the player, year and card number, and eBay sold data gives a value. Around that sits real collection software: infinitely nestable folders, checklist tracking that auto-fills from your collection as you complete sets, community checklist templates, multi-attribute sorting, iCloud sync, CSV export and image sharing. Version 4.3 added full database search, so you can type "2011 Mike Trout #US185" instead of photographing it. Sources: https://cardstockapp.com and the App Store listing https://apps.apple.com/us/app/cardstock-sports-card-scanner/id1508888618, read 2026-09-16.

Harmonic resolves cardstockapp.com to Open Brace Apps, described there as an iOS development shop that also builds for clients. That description is stale: openbraceapps.com now serves the Cardstock marketing site, so the studio has consolidated on its one real product.

Founders and team

Kyle Fowler, owner since 2020-03 (Harmonic person 112753985, https://linkedin.com/in/kylefowler2). He finished St. Dominic High School in 2020 and took a Computer Science bachelor's at Missouri University of Science and Technology from 2020 to 2024. He started Open Brace in March 2020 and shipped Cardstock in April 2020, which means he built and launched it at eighteen, then ran it through four years of an engineering degree. Harmonic lists no other employer in his history. He is unambiguously technical and the app's bundle id, com.KyleFowler.Cardstock, says he wrote it himself.

Jacob Narayan, Chief Operating Officer since 2024-06 (Harmonic person 183548079). Biomedical engineering degree from North Carolina State, 2019 to 2023. His prior and, on Harmonic's record, concurrent employment is a Wendy's crew member role starting 2023-08 with no end date, plus a ski shop technician job and an Amazon delivery contract. A COO title sitting on top of that is a title, not a job.

Sam Aven, software developer since 2025-06 (Harmonic person 185251261). University of Kansas, president of the KU Roundnet club 2024 to 2025, two summer internships at PRI Global. That club presidency explains the studio's Roundnet Score Tracker app.

So: one genuine technical founder in his mid-twenties and two recent graduates. This is a young, informal, bootstrapped shop, and the titles are aspirational.

The studio sprays. Under the Open Brace seller name on 2026-09-16: Cardstock (11,304 ratings), Scanémon trading card scanner (485, launched 2025-07), Brush Buddy toothbrushing (71, launched 2026-03), Costumize digital inventory (19), MagiSnap Magic the Gathering scanner (5, launched 2026-07) and Roundnet Score Tracker (0). Five side products while the one asset that works is losing ground.

Checks run: Harmonic company enrich on cardstockapp.com, Harmonic person records for all three named people, the company site, the App Store listing and in-app purchase table, the App Store most-recent review feed, and the full Open Brace seller catalogue. No adverse record found in the checks run: no litigation, dissolved entity or disputed claim surfaced. The Apify LinkedIn actor was not run because the Harmonic person records already carried degrees, dates and LinkedIn URLs.

Product and traction

11,304 iOS ratings at 4.53 over six years and five months, and Harmonic web traffic of 10,250. That is by far the strongest traction in this batch and it is real: it beats Card Ladder's 9,602, which raised money.

But it is fourth in its own vertical and the gap is widening. On 2026-09-16 the sports card scanner shelf reads CollX 48,085 ratings, Sports Card Investor 37,718, Ludex 27,430, Cardstock 11,304, Card Ladder 9,602. CollX has raised $28.2M and Collectr gives the product away. Cardstock is competing on nothing but the founder's own time.

The current review feed is the alarming part. Pulling most-recent reviews on 2026-09-16 returns, in order: 1 star ("gets the version of almost every card insert wrong", 2026-09-06), 1 star ("gets it wrong 75% of the time", 2026-09-03), 5 star, 3 star ("Right now it does not work just says na for everything also used to be good", 2026-08-30), 1 star, 1 star ("Scanned 10 cards got each of them wrong", 2026-08-26), 1 star ("Said it would give a free trial but still charged $107", 2026-08-24). The 4.53 lifetime average is six years of goodwill masking a product that is currently broken. "Used to be good" and "just says na for everything" describe a comps pipeline that has stopped returning data, which is exactly the platform-dependence failure this category keeps suffering.

No Google Play listing is visible from the site.

Business model and unit economics

Free download with subscription unlocks. In-app purchases on 2026-09-16: $4.99, $14.99, $19.99 and $99.99 tiers labelled "Unlimited Access", plus a $19.99 annual option. The $99.99 tier matches the "still charged $107" complaint, which with tax is the annual plan converting off a trial.

With no outside capital, six years of operation and two junior salaries at most, the business is almost certainly cash positive. That is a real achievement and it is also the whole story: this is a profitable small software business, not a compounding one. There is no take rate, no marketplace integration, no participation in the value of the cards it prices. Revenue is capped at what collectors will pay for a catalogue tool, and the free alternative in the same vertical already exists.

Position in this market

The brief's one constructive finding is that where this thesis compounded, it was vertical with liquid comps, and sports cards is precisely that vertical. Cardstock is in the right lane. The problem is that the right lane is already occupied by companies with twenty-eight million dollars and by a free competitor, and Cardstock's differentiator is collection-management depth rather than recognition accuracy, while recognition accuracy is what is currently failing.

It does not list, cross-list or transact, so it is not exposed to the cross-lister bans. It is also not exposed to any upside. It is not a threat to Moe, whose customer is a consumer selling household goods, not a set collector.

Base decomposition

• Market and competition: 4. The right vertical, with liquid comps and real collector willingness to pay, but Cardstock is fourth in it behind a $28.2M-funded leader and a free product.
• Product and wedge: 4. Genuine software depth in folders, checklists, sync, export and database search that a wrapper cannot copy quickly. Undercut by the core recognition being wrong often enough that recent reviewers say so repeatedly.
• Traction: 5. 11,304 ratings and 10,250 monthly web visits, the strongest in this batch by an order of magnitude and comparable to funded peers. Held at 5 because it took six years and the field has pulled away.
• Unit economics and business model: 4. Almost certainly profitable with no capital and no payroll to speak of, but capped at subscription revenue with no transaction participation, and with trial-conversion complaints that signal non-compounding revenue.
• Team: 3. One real technical founder who built this at eighteen and sustained it for six years, which counts. Around him, a COO whose listed concurrent job is fast food and a 2025 graduate, and a studio splitting attention across five unrelated apps.
• Founder trust and disclosure: 3. Named people with checkable degrees and a real entity. Against that: no pricing page on the site (/pricing 404s), multiple 2026 reviews describing a charge after a stated free trial, and marketing copy promising "precision" and "spot on" scanning that the current review feed contradicts.

Mean = (4 + 4 + 5 + 4 + 3 + 3) / 6 = 23 / 6 = 3.83. Adjusted up 0.17 to 4.00. Six years of survival and eleven thousand ratings with zero outside capital is a harder proof of demand than the component scores capture, and this is the only company in the batch holding a defensible vertical position.

The formula

Base 4.00 + EV 1.00 (Harmonic confirms $0 raised across six years) + Geo 1.00 (St. Peters, Missouri) + Chain 0.00 = COMPANY 6.00.

For Daxos

Pass, and it is the closest pass in the batch. There is no deal on the table: Harmonic's currently_raising is null, the founder has taken nothing in six years, and someone who bootstrapped from eighteen through a degree has revealed a preference for owning the whole thing.

If he did raise, the price would be low, probably $2M to $5M post, and Daxos could write $100k to $250k comfortably. The reason not to is structural. Cardstock is fourth in a vertical where the leader has $28.2M and a serious competitor is free, its recognition accuracy is failing right now in public, and even the winners in this lane are tools rather than demand networks. A cheap entry into fourth place in a capped category does not return a fund.

The founder is worth knowing. A twenty-four year old who shipped and sustained an eleven-thousand-rating product alone is a better bet on his next company than on this one. If he ever raises, take the call. On Cardstock as it stands, DAXOS 3.0. VERDICT PASS.

What could not be checked

Revenue, subscriber count, churn and trial conversion are private and no pricing page exists on the site to check the App Store tiers against. The cause of the current scanning failures could not be established: whether the eBay comps source broke, a model was swapped, or the card database went stale. Google Play presence was not confirmed either way. Whether Open Brace still does client work, as Harmonic's stale description claims, could not be verified now that the studio site redirects to Cardstock.

back to the ranked table

Fifteen years, $2.3M raised and three employees to build a home inventory and maintenance platform whose mobile app has not shipped since 2022 and holds 44 ratings at 2.86, the clearest zombie in the adjacent set.

What it is HomeZada is a web and mobile platform for managing a home across its life cycle: inventory, maintenance, remodel projects, finances, documents, and a 2026 layer of "Homeowner AI" and "Visual Design AI" (homezada.com, 2026-09-16). The inventory module is the piece adjacent to this set: photograph possessions, record value, purchase date, brand, serial number and warranty, attach receipts, export an inventory PDF for an insurance agent (App Store, 2026-09-16). Same capture-and-value motion as a resale app, aimed at insurance adequacy, not a sale. No listing, marketplace or take rate. Scope check passed: funding_total $2,300,000, under the $10.9M threshold, private, not acquired.

Founders and team Three co-founders out of the same prior company, the strongest thing on this card. Person records pulled from Harmonic 2026-09-16.

John Bodrozic, co-founder since 2011-02. BS Mechanical Engineering, Notre Dame. President of Meridian Project Systems 1994-2006, then President of Meridian Systems and GM at Trimble 2006-2010 (person 73258741). Trimble acquired Meridian, so this is a real exit, not a claimed one.

David Ing, co-founder and CTO since 2011-04. BSc Hons Computer Science, Portsmouth. CTO of Meridian Systems 2001-2005, CEO of IngBox Software, co-founder and CTO of Terazen Technology 2007-2011, engineer at Northrop Grumman, architect at Yokogawa Marex (person 86291). Deeply technical, the named CTO.

Elizabeth Dodson, co-founder, MBA Marketing, Loyola Maryland. Director of Business Development at Meridian Systems 1998-2010 (person 160920151).

A credible, senior, previously successful team that spent fifteen years and $2.3M to reach three employees. The founders are three decades into their careers, which matters more to a seed investor than pedigree.

Checks run: Harmonic company and founder records, the site and pricing pages, App Store lookup, web search. No adverse record found. LinkedIn scraping not run.

Product and traction HomeZada Mobile, US App Store: 44 ratings at 2.86, released 2011-11-22, last updated 2022-05-27, version 2.4.0, release note "Update for app links" (iTunes lookup 473722482, 2026-09-16). No real release in four years, rated below three stars. For a product whose core job is photographing objects around a house, an abandoned camera app is the product failing at its central task. The site's 2026 AI marketing describes what the mobile client cannot deliver.

Other signals: headcount 3, LinkedIn followers 1,510, Twitter followers 12,775 (Harmonic, 2026-09-16). Harmonic's valuation field is $11,000,000 marked ESTIMATED, off a $2.1M seed on 2014-10-21 at 20 percent assumed dilution, flagged stale. On a three-person company whose last event was a $250K grant that is arithmetic on an eleven-year-old round, not a price.

Business model and unit economics Essentials free. Premium $99 per year or $15.95 monthly, adding maintenance, projects, finances and reports. Deluxe $189 per year for up to three properties with the AI features, plus $99 per extra property (homezada.com/homeowners/pricing, 2026-09-16). A professional channel aimed at insurance carriers, agents and builders is the only route to scale, since consumer acquisition here is expensive and retention worse: homeowners build an inventory once and never reopen the app. Fifteen years in, that channel has produced three employees.

Position in this market Home inventory is its own graveyard, parallel to the one in the brief. Consumers do not maintain inventories, insurers do not pay them to, and the paying demand sits on the commercial side where contractor software and carrier tooling operate. HomeZada is the consumer-side attempt, alive but not growing for fifteen years. No threat to a resale product: it never touches a marketplace, never generates a listing, and sells a subscription to a filing cabinet, not a transaction.

Base decomposition Market and competition 3.5: home inventory has never produced a venture outcome, consumer retention is poor, and the paying side is commercial. Product and wedge 4.0: broad feature set with a new AI layer, undermined by a mobile client abandoned since 2022 in a fundamentally mobile job. Traction 2.5: 44 iOS ratings in fifteen years, three employees, no disclosed users or revenue, no growth signal. Unit economics 4.5: $99 to $189 per year is a sensible price with a sensible professional channel, but fifteen years of it funded three salaries. Team 6.0: three senior co-founders out of Meridian Systems, a genuine Trimble exit, a thirty-year technical CTO. Founder trust 5.5: long verifiable public record, no adverse findings, but the site markets AI well ahead of the shipped product. Mean of 3.5, 4.0, 2.5, 4.5, 6.0, 5.5 is 26.0 divided by 6 equals 4.33. Adjusted down 0.25 for the four-year-stale mobile app in a mobile-first category and a fifteen-year clock ending in a grant. Base 4.10.

The formula 4.10 + 0.75 (EV, $2.3M raised, $1M to $3M tier) + 1.00 (USA) + 0.00 (chain) = 5.85.

For Daxos, price and verdict DAXOS 1.0. There is no seed to enter. The last capital event was a $250K grant in November 2022, the company has three employees after fifteen years, and the only price marker is an $11M estimate off a 2014 round no buyer would honor. Against the live benchmark here, SellRaze at $500k on a roughly $2.0M Harmonic-estimated post, HomeZada would have to price at a fraction of that stale $11M, and even then the category has no fund-returning ceiling. Daxos brings nothing three experienced operators have missed since 2011. PASS.

What could not be checked Revenue, subscriber count, the size of the professional and insurance channel, Android and web usage, whether the headcount excludes contractors, and the 2022 grant terms. No web funding figure was substituted for the Harmonic record.

back to the ranked table

Free crosslister used as bait for a pallet liquidation marketplace with a real take rate, run by five people on $800k with a Russian state-linked fund on the cap table.

What it is

Closo runs three things on one site: a wholesale marketplace where resellers buy manifested liquidation pallets, a free crosslister that posts one listing to Poshmark, eBay, Mercari, Depop, Vinted and Shopify, and a set of add-ons (Closo Direct storefronts, "six AI agents", Closo Capital working capital, auto-pricer, expenses and taxes). The pricing page title lists tiers as free crosslisting, Member, Executive and Pro Trader (closo.co/pages/pricing, 2026-09-16). Harmonic's description is narrower and older: routing brand returns and excess inventory to professional resellers, replacing the pallet auction channel.

The structure matters. Closo does not try to sell the crosslister. It gives it away and monetizes the sourcing transaction: 0% on buyers the liquidator brings, 10% on buyers Closo brings, plus 3.9% and 30 cents on payments (closo.co, 2026-09-16).

Founders and team

One founder is visible: Ivan Cerkasin, who also appears publicly as Ivan Cherky. Both LinkedIn URLs carry the same numeric suffix (b99518129), so this is one renamed profile, not two people. Prior roles surfaced in search are MyTable and a family furniture factory (ZoomInfo profile, accessed 2026-09-16). No engineering degree, shipped code or GitHub presence turned up, so I record him as non-technical-unknown rather than technical. No CTO or lead engineer is named anywhere I could find. The site has no team page at all.

Checks run: Harmonic person_relationships (returned zero people), company site about and pricing pages, web search on both name spellings, Crunchbase and CB Insights people pages via search results. No adverse record found in the checks run. I did not run LinkedIn scraping.

One disclosure item is real and unflagged by the company. Harmonic lists Internet Initiatives Development Fund (IIDF) as an investor. IIDF is the Russian state-linked venture fund created out of the Agency for Strategic Initiatives. AngelsDeck is a Russian-diaspora angel syndicate. For a US fund this is not a rating question, it is a legal and banking diligence question, and Closo publishes nothing about it.

Product and traction

All traction numbers come from Closo's own marketing pages (2026-09-16) and none are third-party verifiable: 20,000+ resellers, 3,290,000+ listings crosslisted, 33,000 to 35,000 live wholesale lots, 150+ vetted suppliers, six marketplaces. There is no iOS app under this name in the App Store, so the rating-count proxy that works elsewhere in this market is unavailable here. Harmonic headcount is 5.

Business model and unit economics

Take rate on wholesale plus payments, not a tool subscription. That is the right side of the line the category brief draws. The weakness is specific to pallet marketplaces: margins are thin, fraud and trust costs are heavy, and "0% on your buyers" invites the supplier and the reseller to transact off-platform as soon as they know each other. B-Stock, Liquidation.com, BULQ and Direct Liquidation hold the supply relationships and are far better capitalized.

Position in this market

Closo is the only company in this batch that treats crosslisting as customer acquisition rather than product. That is the Croissant-shaped move the brief identifies: get upstream, monetize the transaction. But the demand network it is building is professional resellers buying pallets, a market with entrenched incumbents, not consumer demand.

Base decomposition

• Market and competition 3: liquidation is large but B-Stock and Liquidation.com own supply; the crosslister leg is a free commodity per the brief.
• Product and wedge 5: give away the tool, charge on sourcing. Structurally the best idea in this batch.
• Traction 3: every number is self-reported on its own site, no app ratings, no revenue, 5 headcount.
• Unit economics 4: real 10% plus payments take rate, but thin-margin, high-trust, disintermediation-prone.
• Team 3: five people, one visible founder, no technical lead named, prior work not marketplace or logistics at scale.
• Founder trust and disclosure 3: two published name spellings, conflicting founding dates (Harmonic 2020 vs profiles 2023), no team page, undisclosed Russian state-linked investor.

Mean (3+5+3+4+3+3)/6 = 3.50. Adjust +0.25 because the monetization shape is genuinely right. Base 3.75.

Formula

3.75 + 1.00 + 1.00 + 0.00 = 5.75

For Daxos

DAXOS 3.5. No evidence Closo is raising; last round closed 2025-05-28. Entry price unknown, and against the live benchmark in this wedge a $800k-raised company would be pitching somewhere in the low single-digit millions post. The blocker is not price, it is that a $100k to $500k check here buys a five-person team fighting B-Stock on supply, with a cap table that would take real outside counsel time to clear for a US fund, and with every headline metric unverifiable. MEET, not a check: worth one call to ask what the actual wholesale GMV and take-rate revenue are, and to get a straight answer on the IIDF position. If GMV is real and IIDF is cleanly exited, reopen.

VERDICT MEET.

What could not be checked

Actual GMV, revenue, paying seller count, Chrome extension install count, Google Play presence, the precise structure and current status of the IIDF holding, and whether the crosslister uses official APIs or browser automation. The pricing page did not render tier prices to a plain fetch.

back to the ranked table

B2B listing and pricing software for charity and second-hand shops in Europe, the rare company in this set selling to an institution with a budget rather than to a consumer, but capped by a small TAM and a European impact round it just closed.

What it is

Minimist sells an app to charity retailers and second-hand stores. Staff or volunteers photograph an item, the app returns a title, description, attributes, a price range drawn from recent comparable sales, automated garment measurements and an AI-rendered model shot, then publishes to eBay, Shopify and Ricardo (minimist.com/en and /en/pricing, fetched 2026-09-16). There is a second mode that skips online selling entirely and just prints in-store price tags, which is the part no consumer tool does.

The buyer is an organisation, not a reseller. Named customers are Helen and Douglas House (UK hospice), VinoKilo (DE/CH) and Diakonia (DE) (eu-startups.com, 2026-03). The site claims "30+ charity retailers worldwide".

Founders and team

Stephan Hofmann, co-founder and CEO, was a product manager on Google Maps micromobility and on Spotify wearables, with a B.Com in information systems from the University of Canterbury (at.linkedin.com/in/stephanhofmann1, via search 2026-09-16). Robert Andrei Damian, co-founder and CTO, led AI at Panprices, a Swedish price-comparison company. Anna Greil is co-founder and COO. The about page lists nine more names including a full stack engineer (Tomm Jobit), a mobile developer (Mihai Gheorghe), a UK sales executive and an Australia country lead, matching Harmonic's headcount of 10.

Checks run: Harmonic company record (no person_relationships returned), the company about page, web search on both named founders, and a read of the March 2026 EU-Startups round coverage. No adverse record found in the checks run. No litigation, dissolved entity or disputed claim surfaced. I did not run the Apify LinkedIn scraper here because the public profile summary already gave the prior roles. One small inconsistency worth noting rather than treating as adverse: Harmonic dates founding to 2023, the site says 2024.

Product and traction

Reported traction as of the March 2026 round: 150,000 listings created by customers, 17,000 hours of manual work saved, 15 online customers onboarded (eu-startups.com, 2026-03). The iOS app has 0 ratings despite being live since October 2024, but it was last updated 2026-09-13, so it is maintained. Ratings are the wrong proxy for a B2B tool that staff are told to install, so I do not read the zero as death; I read it as no consumer pull at all.

Pricing is tiered by listings per month across Basic, Standard, Pro and Enterprise, in EUR, GBP and AUD, with a free starter tier and a 15 percent annual discount. No price points are published on the page; everything routes to a demo booking.

Business model and unit economics

Per-seat or per-listing SaaS with a hands-on onboarding motion ("our team works closely with yours"). That is services-heavy and will hold gross margin down early, but charity retail is a low-churn buyer once installed and the shops have real budget lines. No ARR is disclosed anywhere I could find. Ten people on $1.52M raised is a short runway; the March 2026 round is roughly six months old as of this write-up.

Position in this market

Minimist is the only company in this batch that avoids the two things that killed the peer set. It sells to a business, so it is not competing with free eBay photo-to-listing aimed at consumers, and it integrates through official APIs (eBay, Shopify, Ricardo) rather than browser-automating Facebook Marketplace, Poshmark or Mercari, so the enforcement history in this lane does not apply to it. The cost is a small TAM: UK charity retail is roughly 11,000 shops, and European second-hand retail chains are a countable market. Even strong penetration at a few hundred euro a month is a single-digit-millions ARR business. The incumbent competition is charity retail EPOS (Cybertill, Nisyst, Eproductive), not the cross-listers.

Base decomposition

• Market and competition 4.0. Real named buyer with budget, but a countable shop-level TAM and entrenched EPOS incumbents.
• Product and wedge 5.5. In-store pricing mode, price tags, virtual models and automated sizing are genuinely not what the platforms give away free.
• Traction 3.5. 150,000 listings and 30-plus retailers are real, but only 15 online customers and no disclosed revenue.
• Unit economics 4.0. Tiered SaaS is the right shape, undermined by high-touch onboarding and no published price or ARR.
• Team 5.0. Technical CTO with applied AI experience, ex-Google and Spotify CEO, ten people shipping across three regions.
• Founder trust and disclosure 5.5. Publishes specific numbers and names customers; Harmonic's $1.52M reconciles with the reported €350k plus €1M.

Mean = (4.0 + 5.5 + 3.5 + 4.0 + 5.0 + 5.5) / 6 = 27.5 / 6 = 4.58. Adjusted +0.17 because it sidesteps the platform-dependence failure mode that killed most of this lane. Base = 4.75.

Formula

4.75 Base + 0.75 EV ($1,520,242, the $1M to $3M tier) + 0.25 Geo (Austria, EU) + 0.00 Chain = COMPANY 5.75.

For Daxos

Not investable for this fund, at any plausible price. Minimist closed a €1M pre-seed in March 2026 led by Tilia Impact Ventures with InvestEU support, so it is not raising and the cap table is European impact capital that a US angel fund cannot easily join at seed. Even if a $100k to $500k allocation appeared, the exit shape is a charity retail software business bought by an EPOS vendor or a PE rollup at a few times ARR, which cannot return a fund. Daxos adds nothing in Austrian charity retail. DAXOS 3.0, VERDICT PASS. It is the strongest business in this batch and still a pass.

What could not be checked

No revenue, ARR, gross margin, churn or paying-customer count is public. Harmonic returned no valuation field and no person records, so the post-money on the March 2026 round is unknown. Google Play presence was not visible from the site pages fetched. I did not verify the LinkedIn profiles directly through the Apify actor.

back to the ranked table

A well-built, honestly presented, entirely bootstrapped solo home inventory app that shipped a release the day this card was written, and is a lifestyle business rather than a venture one.

What it is

Everspruce is a home inventory app on iOS, Android and a new web version. Photograph an item, name it, save it. Free for the first 50 entries, subscription beyond. Features include batch entry, cloud backup, multi-user sharing up to four people, printable QR labels, PDF and CSV export, tags, search, and image import (everspruceapp.com, 2026-09-16). The entity is Lucky Star Labs, LLC.

The About page names resellers as one of five target users, alongside collectors, declutterers, movers and insurance inventories. That is the only thread to Moe's market, and it is thin: Everspruce catalogs, it does not value or list.

Founders and team

Carol George, sole founder, behind Lucky Star Labs, LLC. Her own bio: over nine years of finance and management experience in ecommerce, time off to raise her family, then managing a personal real estate portfolio until 2020. In 2011 she launched Tullie Girl LLC, a girls' accessory online store (everspruceapp.com/about-us, 2026-09-16).

She is not technical. No engineering degree, shipped-code history or GitHub presence appears in the record, and no CTO, engineer or other team member is named on the site, in Harmonic or in search. Everspruce was almost certainly built by contract developers under her direction, the normal path for a non-technical solo founder. Harmonic's record is a stub: no founding date, location, headcount, people or rounds, stage VENTURE_UNKNOWN, funding_total 0. Social handles and nothing else, consistent with a company that has never raised.

Checks run: Harmonic company record, App Store search and lookup including seller entity and price, direct fetch of the home, about and pricing pages, web search on the founder and the LLC. A BBB profile and a Google Play developer page both exist for Lucky Star Labs, LLC. No adverse record found in the checks run.

Product and traction

US App Store, 2026-09-16: Everspruce Home Inventory, 473 ratings at 4.66 average, released 2018-08-02, version 1.28.0, updated 2026-09-16, Free with subscription. It was updated the same day this card was written, and the site banner promotes a newly launched web app. For a bootstrapped solo operation eight years in, that cadence is the most credible thing in the file.

Self-reported site metrics: 20,000 installs worldwide, 1,100,000 entries added, 4.7 average US rating. Modest and specific enough to be plausible, and the App Store average of 4.66 corroborates the rating claim. Instagram 5,609, Facebook 200, Twitter 333. Harmonic has no web traffic figure. 473 ratings in 97 months is roughly five a month: better per year than Itemtopia or Nest Egg, an order of magnitude below Dupe.com.

Business model and unit economics

Freemium subscription: 50 entries free forever, paid beyond. The correct model, and the specific thing Nest Egg got wrong with its $6.99 one-time purchase. There is no pricing page; /pricing returns 404, so the price is visible only inside the App Store purchase flow, a real disclosure gap for a product asking people to subscribe. With zero raised and no payroll, the cost base is small enough that 20,000 installs converting at a low single-digit rate could cover it. A plausible profitable micro-business, not a plausible venture return.

Position in this market

Everspruce sits in the consumer home-inventory wedge with Itemtopia and Nest Egg, the one Sortly abandoned fourteen years ago. Cataloging possessions is free inside Google Lens, Apple Visual Intelligence and every marketplace app, and consumers have never paid much for the standalone version. Everspruce's response has been to stay small, take no outside money and ship steadily: the rational strategy, and the reason it will never be a venture outcome.

Base decomposition

• Market and competition: 3.5. A wedge with a long documented record of not converting, now commoditized at the OS layer.
• Product and wedge: 4.5. A good feature set for a solo build: batch mode, QR labels, multi-user, exports, tags, three platforms, a new web app.
• Traction: 2.5. 473 ratings over 97 months, 20,000 self-reported installs, no independent usage data.
• Unit economics: 4. Freemium subscription is the right model, and zero capital consumed with no payroll means it can be profitable at small scale. Price not disclosed publicly.
• Team: 3. Solo, non-technical, first software product, no engineer named. Eight years of steady shipping is the credit.
• Founder trust and disclosure: 4.5. A modest, checkable story with no inflated claims, self-reported metrics corroborated where possible, nothing adverse. The missing pricing page is the one gap.

Mean = (3.5 + 4.5 + 2.5 + 4 + 3 + 4.5) / 6 = 22 / 6 = 3.67. No adjustment. Base 3.67.

Formula

3.67 Base + 1.00 EV ($0 raised, the $0-$1M tier) + 1.00 Geo (USA) + 0.00 Chain = COMPANY 5.67.

For Daxos

PASS. Harmonic shows zero rounds, zero dollars raised and currently_raising null, so no valuation figure exists. The entity is an LLC, not a C-corp, so a priced round would require conversion first, and nothing suggests the founder wants that.

She has spent eight years deliberately building a business that does not need capital, in a wedge where capital has repeatedly failed. Taking $100k to $500k would force growth expectations onto a product whose advantage is that it has none. A seed check cannot return a fund here: the ceiling on a home inventory subscription with 20,000 installs is a few hundred thousand dollars of annual revenue, and this category has never produced an acquirer who pays up for one. Daxos adds nothing the founder needs. Fit 1.5 out of 10.

Threat to Moe: 1. Catalogs items, does not price or list them.

What could not be checked

Subscription price, subscriber count, revenue, churn and Google Play install count are undisclosed. The 20,000 installs and 1,100,000 entries figures are self-reported with no date and no independent corroboration. Harmonic has no founding date, location, headcount or people, so those rest on the App Store release date and the absence of named staff. Who writes the code could not be established, and whether the LLC took friends-and-family money outside Harmonic's coverage is unknown.

back to the ranked table

Six-week-old YC S26 solo founder selling an AI sales agent into other people's marketplaces, which escapes this category's ceiling by leaving the category.

What it is

Palisade installs a personal sales agent on a marketplace operator's own site. Every visitor gets an agent that knows the catalogue, learns taste across sessions on web, SMS and email, quotes live availability and real prices, and navigates the site with its own cursor from discovery through checkout. Sellers get an agent for performance data, listing rewrites, repricing and underperformance alerts. The operator's team gets high-intent leads in CRM with the conversation attached, plus human handoff (palisade.run, read 2026-09-16). YC Summer 2026. One customer is publicly live: Shorebreak, a vacation rental operator, at shorebreak.palisade.run.

A records discrepancy worth flagging. The batch file recorded "Harmonic: no record" as of 2026-09-15. A fresh pull on 2026-09-16 returns a full record: $500,000 pre-seed dated 2026-07-01, Y Combinator, PRE_SEED, headcount 1, founded 2026, with a $2.0M post marked ESTIMATED.

Founders and team

Jonathan Salama, founder and CEO, and as far as every source shows the entire company. Princeton CS, software engineer at Amazon on Traffic and Marketing, founding engineer at Sage AI Labs, co-creator of Lulus TV, a YouTube channel with 1M subscribers (YC bio, read 2026-09-16).

Clearly technical, and the YouTube channel is a real distribution credential rather than padding. He is also alone: headcount 1, one active founder on the YC page, zero open jobs. A solo founder six weeks past Demo Day selling infrastructure into marketplaces is the central risk here, not the technology.

Checks run: Harmonic enrich, YC company and launch pages, web search on the founder and company, site fetch, iTunes search. No adverse record found in the checks run: no litigation, no dissolved entities, no disputed claims. Naming caution: a different Jonathan Salama is CTO of Transfix, and an unrelated Palisade in the same batch does OS vulnerability remediation (ycombinator.com/companies/palisade-2).

Product and traction

One self-reported number exists: "Our agents have driven over $2.7M in GMV since the start of the batch" (YC launch post, read 2026-09-16). No methodology, no attribution window, no statement of whether that is incremental GMV or all GMV on sites running the agent. Unverified. One named production customer, Shorebreak, on a palisade.run subdomain. The site names trading card platforms and vintage stores as targets but shows no logo for either. No pricing published. No iOS app, correctly, since this is embedded B2B software.

Business model and unit economics

B2B SaaS or usage-based pricing into marketplace operators, undisclosed. This is the best structural feature in the batch: it takes money from businesses rather than consumers, sidestepping the demand-network ceiling that caps every seller tool here. The unproven part is inference cost. An agent running for every visitor, holding memory across sessions and driving a cursor, burns tokens per session against a contract priced per month. Nothing published shows that spread.

Position in this market

Adjacent, and only barely. Palisade is not in the consumer scan, value and list loop, does not touch cross-listing, and would sooner sell to a resale marketplace than compete with one. Threat to Moe: 1. It is an AI-agents-for-ecommerce company that names marketplaces as its wedge, which puts it against Shopify's agents, Rep AI, Zoovu, Algolia and every platform-native shopping assistant, a lane more crowded and better funded than resale tooling.

Base decomposition

• Market and competition: 3.0. Large TAM, but the AI sales agent lane is saturated and platforms ship this natively.
• Product and wedge: 4.0. An agent taking real actions on the merchant's own UI, plus the three-sided framing, is differentiated.
• Traction: 3.5. One live logo, one self-reported GMV figure, six weeks old.
• Unit economics: 3.5. B2B is the right side of the table; agent cost per visitor is unproven.
• Team: 4.0. Strong technical solo founder with real distribution history, but headcount 1.
• Founder trust: 3.5. The $2.7M claim carries no methodology. Nothing adverse found.

Mean = (3.0 + 4.0 + 3.5 + 3.5 + 4.0 + 3.5) / 6 = 21.5 / 6 = 3.58. No adjustment; too young for a tilt either way. Base = 3.58.

The formula

3.58 + 1.00 (EV, $500k, the $0 to $1M tier) + 1.00 (Geo, USA) + 0.00 (Chain) = 5.58.

For Daxos, price and verdict

Harmonic's ESTIMATED $2.0M post is the YC standard deal, not a market price. A company six weeks out of an S26 Demo Day with a $2.7M GMV headline will raise at $15M to $25M post, and AI agent companies out of recent batches have cleared higher. At that price Daxos buys well under two percent, from a solo founder, in a lane where it has no customer introductions and no sector edge. The one honest argument for engaging is that Palisade's model escapes the category ceiling, and it escapes it by not being in the category. That is interesting reading, not a check. PASS at any post-Demo-Day price. MEET only if the round prices near the Harmonic estimate, which it will not, or if a second technical founder joins and two named resale logos go live with disclosed contract values.

What could not be checked

Pricing, contract values, customer count beyond the one visible logo. Whether the $2.7M GMV is incremental or total. Gross margin after inference. Whether a raise is open and at what price. Whether Salama is recruiting a cofounder. No LinkedIn scrape: the Apify actor failed twice earlier in this batch and was stopped per the rules.

back to the ranked table

A vertical collectibles exchange with instant bid/ask and a vault, not a cross-lister, priced at a Harmonic-estimated $23M post that a seed check cannot clear.

What it is

Tradepost (legal name Tradepost Markets, Inc.) runs a mobile exchange for trading cards, sports cards, coins, LEGO sets and live event tickets. The pitch on the site is "buy lows, sell highs" against inventory that is verified and held in a Tradepost vault, plus live trading and a holdings view showing live market value (tradepost.co, fetched 2026-09-16). Harmonic's own summary is "sell anything in 30 seconds," which is the earlier and broader framing; the shipped product is narrow and vertical.

This matters for placement. Tradepost is not in the seller-tool lane at all. It is a demand network with liquid comps, the shape the category brief identifies as the only one that has compounded. It is closer to Alt or CollX than to a cross-lister.

Founders and team

Tyler Mozeleski, co-founder and CEO since 2024-08. BS Economics, Moravian University, 2019 to 2023. Machine Learning Engineer at Alt from 2023-09 to 2024-08, and before that a Goldman Sachs global markets summer analyst in 2022. Founded Cardboard Investment Group LLC in 2020 while an undergraduate. Alt is the most relevant possible prior employer for this product and he was on the data side of it (Harmonic person 153008493).

Alexander Perez, co-founder and CTO. BS Computer Engineering, Syracuse, 2019 to 2023. Three software engineering internships at SAS, then a year as a Goldman Sachs data analyst 2023-07 to 2024-07. Joined as CTO of the Backstage build in 2023-12, current title since 2025-04 (Harmonic person 85026169). Technical founder confirmed by degree and engineering employment.

Ryan Bella, original co-founder from 2022-11, left 2025-06 and is now a financial analyst at Vivo Infusion. Prior roles were finance, at Lockheed Martin. Two founding engineers joined 2025-06: Mehtaab Gill (prior engineering at Lewk, EDGE Finance, eStock) and Parker Stevenson (CS, Oregon; a self-employed reseller from 2017 to 2024, which is real domain exposure).

Checks run: Harmonic company record and all six Harmonic person records, company site, App Store listing and description, Tradepost fee schedule. No adverse record found in the checks run. The company has been through at least one name and product change (the Backstage titles in the employment history), and one of three co-founders has exited to an unrelated salaried job, which is worth asking about.

Product and traction

iOS app "Tradepost: Trade Culture," id 6753876601, released 2025-11-12, at version 3.0.0, last updated 2026-09-14, 254 ratings and a 4.68 average as of 2026-09-16 (itunes.apple.com/lookup, 2026-09-16). Android is linked from the site. Harmonic web traffic 16,120. X account 11,058 followers, LinkedIn 784.

Set that against the adjacent benchmark in this market: Collectr sits at 64,932 ratings while charging nothing. Tradepost has roughly one four-hundredth of that after ten months of App Store life, and it launched a full year after taking its first money.

Business model and unit economics

The fee schedule, revised 2026-09-14, is explicit: prices are all-in, Tradepost's compensation "is included within these all-in figures. It is not a separately stated rate, it may vary from transaction to transaction, and it may change at any time without advance notice" (tradepost.co/fees, 2026-09-16). That is a spread, not a take rate. Combined with the vault, it means Tradepost is warehousing collectible inventory on a $4.9M balance sheet and earning the bid-ask. The margin is undisclosed and the inventory risk is real. A market maker in an illiquid asset class needs either much more capital or much tighter turns than this cap table supports.

Position in this market

Tradepost is only a side-swipe at the scan-and-list wedge. It touches the same intent, sell my stuff quickly, but only inside collectibles, and it solves it with a balance sheet rather than a listing tool. Its actual competitors are Whatnot at a $20B valuation, eBay through TCGplayer, Fanatics Collect, and Alt at $346M. Those are the wrong opponents to meet with five people and $4.9M.

Base decomposition

Market and competition 4.5: liquid comps and a genuine demand network, but the incumbents are enormous and well funded. Product and wedge 5.5: instant two-sided liquidity plus vaulting is a real product, not a wrapper on a platform API. Traction 4.0: 254 ratings in ten months, weak against the free adjacent comps. Unit economics 3.5: undisclosed spread, inventory on balance sheet, capital intensive. Team 5.0: relevant technical founders, ex-Alt CEO, but very young and one co-founder already gone. Founder trust and disclosure 3.5: a16z is claimed in a LinkedIn headline but no investor appears in Harmonic, and the fee schedule deliberately hides the take.

Mean of 4.5, 5.5, 4.0, 3.5, 5.0, 3.5 is 4.33. Adjusted down 0.25 for the combination of an undisclosed take rate and co-founder churn at a company with 254 ratings. Base 4.08.

Formula

4.08 base + 0.50 EV + 1.00 geo + 0.00 chain = 5.58 company.

For Daxos, price and verdict

Harmonic shows the seed closed 2025-09-24 at $4.65M with an estimated $23M post (ESTIMATED, derived from assumed 20 percent dilution, so treat it as a range not a fact). currently_raising is null. Even at half that estimate the entry is an order of magnitude above the $2.0M post that this wedge has actually cleared. A $100k to $500k check at $23M buys about 2 percent of a capital-intensive market maker that must beat Whatnot and eBay to matter, and Daxos adds nothing to a collectibles exchange. Daxos fit 3.0. PASS.

What could not be checked

GMV, spread, vault inventory value, Android install count, cohort retention, whether the a16z claim is a direct check or an accelerator program, and the terms of the September 2025 round beyond Harmonic's estimate.

back to the ranked table

A twenty-three app scan-and-value factory with 728,785 US ratings that already shipped a thrift-flipping app into this exact wedge, watched it stall at 49 ratings, and can relaunch whenever it wants.

What it is

Next Vision Limited is an app studio, not a startup. It runs 23 apps on the US App Store under one seller account (artistId 1532006906), almost all of them a camera pointed at a thing that returns an identification and, increasingly, a value. Coins, banknotes, stamps, comics, vinyl, TCG cards, sports cards, antiques, rocks, mushrooms, birds, insects, fish, plus a few fitness and calorie apps.

The five apps named in this row are the collectibles-and-resale cluster: CoinSnap (2022-08), FoilSnap (2025-09), AntiqSnap (2025-10), StarSnap (2025-10), and FlipSnap (2026-03), the one that aims straight at thrift flipping.

Harmonic has no company record. Enriching coinidentifierai.com returns a stub named "Coin" with zero funding, zero people, no location and no founding date; enriching nextvisionlimited.com returns "Company not found; scheduled for enrichment." Every number below comes from Apple, not from a funding database, and no funding figure should be inferred from the absence.

Founders and team

No founder, executive or engineer is identifiable. The seller name is a bare "Limited" with no address. The developer site for CoinSnap (coinidentifierai.com, 2026-09-16) has no about page and no team page; the linked terms and privacy pages on app-service.coinidentifierai.com return 403 behind a Cloudflare bot wall, so even the governing-law clause is unreadable. Harmonic carries no person records for the entity.

One observation worth recording, not a verified fact: the portfolio includes Picture Insect, Picture Bird, Picture Mushroom, Picture Fish and Picture Nature, which share naming and product design with the Glority group's Picture family. If that affiliation is real it means a large, long-established Asian app group is behind this, with the balance sheet and user-acquisition machinery that implies. I could not confirm it in the checks run.

Checks run: Harmonic enrich on two candidate domains, iTunes lookup on the seller's full catalogue, iTunes lookup and the public App Store page for CoinSnap and FlipSnap, the CoinSnap developer site, and an attempt at the legal pages which was blocked. No adverse record found in the checks run, though an opaque entity with no named humans is close to uncheckable.

Product and traction

Catalogue totals as of 2026-09-16 (itunes.apple.com lookup, artistId 1532006906): 23 apps, 728,785 US ratings combined.

The relevant five: CoinSnap 302,489 ratings at 4.72, released 2022-08-03, updated 2026-09-08. AntiqSnap 31,338 at 4.68, released 2025-10-15, updated 2026-09-11. StarSnap 26,999 at 4.68, released 2025-10-24, updated 2026-08-14. FoilSnap 13,375 at 4.76, released 2025-09-30, updated 2026-09-10. FlipSnap 49 at 4.71, released 2026-03-02, last updated 2026-07-09.

Two of those numbers matter more than the rest. AntiqSnap took 31,338 ratings in eleven months, which is more than twenty times what any venture-funded app in this batch has managed. And FlipSnap, the one that went at the flipping wedge directly, has 49 ratings after six months and has not shipped an update since July. VinylSnap tells the same story: 3,329 ratings and no update since 2025-11. ToyWorth, launched 2026-06, sits at 655.

So the studio wins overwhelmingly when the job is identify-and-value a collectible with a known catalogue, and it has so far failed when the job is decide-what-to-flip.

Business model and unit economics

Pure consumer subscription behind a scan paywall. Observed price points on the App Store pages are $4.99 and $7.99 weekly and $29.99 and $39.99 annual, with CoinSnap Premium sold as a yearly plan after a seven-day trial (apps.apple.com, 2026-09-16). No marketplace, no take rate, no listing, no inventory.

This model is well understood: buy installs, convert a few percent at a weekly price, live on the gap between lifetime value and acquisition cost. It is genuinely profitable at scale and genuinely fragile, because it depends on paid user acquisition and on weekly subscriptions that churn hard. Nothing here compounds into a network.

Position in this market

Next Vision is not a competitor for capital. It is the answer to a question the category brief already raises, which is who owns the "what is this worth" surface. The answer is not a venture-funded startup. It is a factory that can stand up a new vertical scanner in about six weeks and has done so twelve times in the last eighteen months.

For anyone building in the scan-and-list lane, the practical read is this. The valuation half of the loop is a commodity with a large incumbent distributing it at $39.99 a year, on top of what eBay, Google Lens and Apple Visual Intelligence already give away. The listing half, the account connections, the marketplace policy risk, the post-sale workflow, is the part Next Vision has not built and, judging by FlipSnap, is not good at. That is thin ground to defend but it is the only ground there is.

Base decomposition

Market and competition 5.0: owns real share of a huge shallow surface, in a knife fight with CoinIn, AIBY, 326 LAB, Scaling Engine and a dozen others. Product and wedge 5.0: repeatable and fast to ship, but shallow, with no listing, no account graph and no reason to stay. Traction 7.0: 728,785 US ratings is an order of magnitude beyond anything else in this category sample. Unit economics 5.5: proven paywall subscriptions at $39.99 a year, offset by paid-UA dependence and weekly-plan churn. Team 2.0: no identifiable humans. Founder trust and disclosure 2.5: opaque entity, unreachable legal terms, $7.99-a-week pricing on a consumer utility.

Mean of 5.0, 5.0, 7.0, 5.5, 2.0, 2.5 is 4.50. No adjustment. Base 4.50.

Formula

4.50 base + 1.00 EV + 0.00 geo + 0.00 chain = 5.50 company.

The EV bonus is mechanically correct and economically meaningless here: Harmonic records no raise because this is not a venture-funded entity, not because it is capital efficient. The geo bonus is withheld rather than guessed, since no jurisdiction could be verified.

For Daxos, price and verdict

Uninvestable. There is no round, no named counterparty, no identifiable jurisdiction, and every sign of a self-funded app studio that has no use for $100k to $500k. Daxos fit 0.5. PASS.

The value of this card is comparative, not investable. It is the benchmark that any funded company in this batch should be measured against, and most of them lose by two orders of magnitude on distribution.

What could not be checked

Corporate ownership and any Glority affiliation, jurisdiction, revenue, headcount, funding of any kind, Google Play scale, paid-UA spend, and the governing terms of the apps, which are behind a bot wall.

back to the ranked table

A solo-built, bootstrapped reseller bookkeeping app that deliberately is not a crosslister, profitable-shaped and far too small to be a venture asset.

What it is

Flippd is an inventory, profit, expense and mileage tracker for resellers, on iOS, Android and web. The App Store description states the position plainly: "Flippd is an inventory and profit tracker, not a cross-listing tool" (iTunes lookup id 1633899291, 2026-09-16). The batch file classed it as a cross-lister; that is wrong and the correction changes the analysis.

Workflow per getflippd.com (2026-09-16): research an item against eBay sold comps before buying, log the purchase, snap receipts for expense extraction, mark which platforms it was listed on (manual, not automated posting), record the sale, log sourcing mileage for tax. An overnight "Coach" reads the books and tells the seller what to fix. eBay sync for listing and sales import is live for all subscribers via official integration.

Founders and team

Shane Jordan is the sole founder and builder. The about page states he has been building software for more than 25 to 30 years and works on cloud, AI and machine learning. He is clearly technical: he ships and maintains three clients himself, and his own address appears on the App Store record and his email (shane@getflippd.com) on the site. Entity is Flippd, LLC, registered in Michigan.

Checks run: Harmonic enrich (no people, no location, no founding date, no funding), Harmonic person lookup not possible with zero person ids, company site about and FAQ pages, App Store developer record (id 1633899293), Google Play listing, web search on the founder name. No co-founder, no team, no prior exit and no prior company surfaced. No adverse record found in the checks run.

Product and traction

135 iOS ratings at 4.71 average, 3 years and 8 months after first release, last updated 2026-08-31 (checked 2026-09-16). Also on Google Play. For scale inside this exact market on the same day: Flippr 1,570 ratings after 14 months, ThriftAI 20,384, Swoopa 541, OneShop 467. Flippd is near the bottom of the comparable set on the one traction proxy this market makes available, despite the highest rating quality in the group.

Business model and unit economics

Free tier capped at 20 inventory items, 25 expenses and 30 mileage entries per month. Pro at $9.99 monthly or $99.99 annually, one subscription across iOS, Android and web, up to 5 team seats (getflippd.com/pricing, 2026-09-16). One operator, no payroll, no marketplace automation infrastructure to maintain, one official eBay integration. This is the healthiest real margin structure in the batch and almost certainly cash-positive at a modest scale. At 135 iOS ratings the paying base is plausibly in the hundreds to low thousands, so revenue is likely in the tens to low hundreds of thousands per year.

Position in this market

Flippd sits in the bootstrapped-survivor lane the brief identifies, alongside List Perfectly at $0 raised, Crosslist at $0 and PrimeLister at $0. It also sidesteps the structural risk that killed or stalled the funded companies. Because it never posts to a marketplace on the seller's behalf, Mercari's August 2024 change, the Poshmark suspensions of May and November 2025 and eBay's February 2026 ban on LLM-driven agentic bots do not touch it. That is genuine, and it is also why the product is small: the thing it does is a spreadsheet replacement, and the free competitor is a spreadsheet.

Base decomposition

• Market and competition 3: reseller bookkeeping is real but small and served by My Reseller Genie, Craftybase, platform seller hubs and free spreadsheets.
• Product and wedge 4: clean, honest, differentiated from crosslisters, and immune to platform bans. Nothing defensible about it.
• Traction 2: 135 iOS ratings in 44 months, bottom of the comparable set.
• Unit economics 5: solo operator, near zero burn, $99.99 annual, one official API. Best real margins in the batch.
• Team 3: one technical founder who ships across three platforms; 25 to 30 years experience is his own claim, unverified independently. No co-founder, no prior exit found.
• Founder trust and disclosure 5: real LLC, real address, founder's own email, published pricing and free-tier caps, and an App Store description that talks the product down rather than up.

Mean (3+4+2+5+3+5)/6 = 3.67. Adjust -0.25 because there is no growth vector at all: one person, no raise, no distribution engine, and the traction curve is flat against peers. Base 3.42.

Formula

3.42 + 1.00 + 1.00 + 0.00 = 5.42

For Daxos

DAXOS 1.5. Flippd has raised nothing across three and a half years and gives every sign of not wanting to. There is no round to price, no co-founder, no team to scale and no plausible path from a $99.99 per year bookkeeping app to an outcome that returns any part of a seed fund, in a category whose brief already shows that seller tools have never produced a venture-scale exit. Daxos brings nothing a solo Michigan developer needs. This is a good small software business and a non-investment.

VERDICT PASS.

What could not be checked

Subscriber count, revenue, Google Play install and rating counts, churn, whether Shane Jordan works on this full time, and any prior companies he built. Harmonic holds no funding, location, headcount or founding record for the entity, so everything outside the Harmonic-verified $0 raise comes from the company's own site and the App Store.

back to the ranked table

A one-person, eleven-year-old home inventory app run part-time by a founder who genuinely sold his last asset-tracking company to BlackBerry, with almost no users to show for it.

What it is

Itemtopia is a home inventory app: document rooms, belongings, receipts, warranties and service history, and produce insurance-ready reports. Four use cases (Homes, Business Pro, Properties, Collections), claiming "10+ years, 100+ countries, 11 languages" across iOS, Android, Mac, Windows and a new web version (itemtopia.com, 2026-09-16).

The batch file records Itemtopia as a zombie. The evidence says otherwise: the main app was updated 2026-09-05, Itemtopia Lite shipped 2026-01-13, the site has been rebuilt and a web version is promoted as new. The company is live. It simply has almost no users.

Founders and team

Steve Hulaj, Founder and CEO since 2015 (Harmonic person 188849330). Company headcount is 1, so he is founder, CEO and engineering team. His history is the most interesting thing here, and it checks out. He founded Nexent Innovations in 1994, an asset management, CRM and ERP product that reached 30 countries, spun the mobile patents into Nextair Inc. in 1998, raised $1M for 15 percent, and took it onto the TSX at roughly $55M. Nextair sold item and asset tracking into 45-plus countries and was acquired by BlackBerry in 2005 (Crunchbase person profile, corroborated by company and press bios). He was CTO and VP Operations at Eldorado Logistics 2004-10 to 2005-11. A real prior exit in the same domain, and the strongest founder credential in this batch after Dupe.com's.

The offsetting finding is equally clear. Harmonic records him as CEO of Centennial Developments Corporation from 2006-02 to present, concurrent with Itemtopia for its entire life. Itemtopia is a part-time venture at a headcount of one. A third-party tracker reports $139.1K revenue with a five-person team in 2021 (getlatka.com/companies/itemtopia), self-reported and five years stale, and the team has since gone from five to one.

Checks run: Harmonic company and person records, App Store US and CA listings and seller entity, company site, web search on the founder and his prior companies. No adverse record found in the checks run.

Product and traction

US App Store, 2026-09-16: Itemtopia Inventory, 166 ratings at 4.48, released 2015-01-27, updated 2026-09-05. Canadian store 69 ratings. Itemtopia Lite, released 2026-01-13, 0 ratings, untouched since 2026-01-16. Harmonic web traffic 3,273 monthly. Instagram 2,804, Twitter 483, LinkedIn 10 followers.

166 US ratings in 140 months is about one a month, the profile of a hobby product. The product surface is unusually wide for one person: five platforms, eleven languages, warranty and service tracking, QR labelling, insurance binder export.

Business model and unit economics

Freemium download with a paid Business Pro tier. At a headcount of one the cost base is near zero, which is why an eleven-year-old product with 235 App Store ratings still ships releases in September 2026. It can persist indefinitely. It cannot compound. No pricing or revenue is disclosed.

Position in this market

Itemtopia sits in the home-inventory adjacency, one step from resale, in the wedge Sortly abandoned in 2012 and that Nest Egg and Everspruce also work. Consumers do not pay to catalog their belongings, and the durable version sells to contractors, insurers or estate professionals. Itemtopia has a Business Pro tier gesturing at that, eleven years in, with one employee.

Base decomposition

• Market and competition: 4. Home inventory for insurance is real but low willingness to pay, with an eleven-year record of not working here.
• Product and wedge: 5. Broad and mature for a solo build: five platforms, eleven languages, warranties, service history, insurance export.
• Traction: 2. 235 App Store ratings across both stores in 140 months, 3,273 monthly visits, 10 LinkedIn followers, headcount down from five to one.
• Unit economics: 3.5. Near-zero cost base makes survival cheap, but nothing suggests the paid tier converts at any useful rate.
• Team: 5. A repeat founder with a TSX listing and a BlackBerry sale in this exact domain, running the company alone and part-time alongside a CEO role held since 2006.
• Founder trust and disclosure: 4. Prior-exit claims corroborated rather than inflated. Site stats are unfalsifiable vanity metrics. No adverse record.

Mean = (4 + 5 + 2 + 3.5 + 5 + 4) / 6 = 23.5 / 6 = 3.92. Adjust minus 0.25: eleven years in, the team has shrunk to one, the founder's attention is split with another CEO role, and the January 2026 product has zero ratings and no update since. Base 3.67.

Formula

3.67 Base + 1.00 EV ($1,000,000 exactly, which the boundary rule places in the $0-$1M tier) + 0.75 Geo (Canada) + 0.00 Chain = COMPANY 5.42.

For Daxos

PASS. Harmonic shows currently_raising null, one round closed 2021-01-05, no valuation figure. There is no round to enter and no reason to create one: a $100k to $500k check would fund a one-person, part-time, eleven-year-old consumer app with 235 App Store ratings, in a wedge this category's history says does not convert.

The founder is the only asset worth noting. If Hulaj left Centennial Developments and attacked the insurance or estate channel full time with a real team, that would be a different conversation. Nothing suggests he intends to. Fit 1.5 out of 10.

Threat to Moe: 1. No listing, pricing or marketplace function.

What could not be checked

Revenue, subscriber count and paid conversion are undisclosed; the only figure is a self-reported $139.1K from 2021 on a third-party tracker. The 2021 seed investor is unnamed in Harmonic. Terms of the BlackBerry sale could not be established. Google Play installs were not visible, and why headcount fell from five to one is undocumented.

back to the ranked table

The deepest product in this batch built by one iOS engineer over four years, with garment measurement and ghost-mannequin photography the funded cross-listers do not have, and 126 ratings to show for it because there is no website and no marketing.

What it is

A full reseller workflow in one iOS app: write a listing once and post it to eBay, Mercari, Poshmark, Depop and Grailed; scan an item by camera, barcode or search before buying for average sold price, demand, margin and comps; photograph a garment once and get chest, length and sleeve measurements; remove backgrounds, generate ghost-mannequin shots and place clothing on AI models; scan a care label for brand, size, material and care; track inventory and sales across all five platforms (App Store description, read 2026-09-16).

A cross-lister, sourcing scanner, photo studio and inventory system in one app. The broadest feature set in this batch by a distance.

There is effectively no company around it. trylista.com is a Google Sites page whose only pages are Home, Privacy Policy and Terms. Harmonic's crawler read it and recorded the company description as, literally, "Privacy Policy" (Harmonic, 2026-09-16). The App Store seller of record is a person, Luis Galvan, not an entity.

Founders and team

Luis Galvan, sole developer, Los Angeles County, California. UC Irvine, bachelor's degree, 2016 to 2020. He calls himself "a knowledge seeking, passionate, and fast learning iOS developer" with a good understanding of Swift, and lists among his hobbies "analyzing secondary marketplaces" (LinkedIn via Apify harvestapi, 2026-09-16). Skills: Swift, UIKit, Xcode, iOS design. Genuinely technical, which separates him from most solo builders in this set.

Four apps shipped: Biblo, book trading, 2020-11 to 2023-10; ResellX, an autofill and flip-information tool, 2021-02 to 2023-10; Lista from 2022-02 to present; and a contract as iOS Developer at OnlyChitChat, 2024-10 to 2025-03. ResellX means he has built resale tooling continuously since February 2021, five and a half years on the same problem, with Lista as the second generation. The mid-life contract suggests the app was not paying his bills at that point.

He titles himself "iOS Developer at Lista", not founder. No co-founder and nobody doing distribution, which the traction shows.

Checks run: Harmonic enrichment on trylista.com (stub, no founding date, location, headcount or person relationships); the App Store listing, lookup and web page; the trylista.com pages; a LinkedIn pull via Apify; web search on the developer, app and domain. No adverse record found in the checks run: no litigation, no dissolved entities, no disputed or inflated claims, and education and employment are consistent across every source.

Product and traction

126 ratings at 4.61, released 2022-08-16, updated 2026-09-09 v2.3.45 (iTunes, 2026-09-16). The latest release landed a week before this review, noted as "several marketplace fixes and UI improvements", which is what maintaining unofficial integrations looks like week to week.

126 ratings over forty-nine months is very low. In the same store search Vendoo has 2,784, OneShop 467, Crosslist 293 after five months, ClosetMate 7,337, SellRaze 39,656. Lista has the best product in that list and close to the worst distribution. The reason is visible: no marketing site, no blog, no social presence recorded by Harmonic, no Google Play build found. Everything it gets comes from App Store search.

The listing claims "thousands of resellers"; 126 ratings does not support that and no other source does. Review summaries surfaced in search describe both fast feature turnaround and complaints about support and duplicate listings. That is second-hand and was not verified against the review stream.

Business model and unit economics

Free download with an auto-renewing subscription. The App Store page did not expose tier prices to the fetches run, so the price is unknown.

The cost structure is the worst in this batch. Measurement, background removal, ghost-mannequin generation, model try-on, label parsing and listing copy are all generative calls, so a heavy user costs real money on every item, not once a month. ClosetMate charges $9.99 a month for automation with essentially no inference cost; Lista charges a consumer subscription against per-item generative spend. Whether that clears is the first thing to ask. Offsetting it: no payroll, no burn beyond one person's time and API bills.

Position in this market

The cross-lister lane, which has a hard ceiling. Flyp raised $15.8M, has not priced a round since 2022-05, and gives the product away free, so Lista competes against free. The survivors, List Perfectly, Crosslist and PrimeLister, are all bootstrapped, and Lista fits that pattern.

Platform exposure is total. None of the five marketplaces offers a public consumer cross-listing API for this purpose. Mercari's August 2024 change silently broke cross-listers, Poshmark suspended cross-lister users in May and November 2025, and eBay's February 2026 user agreement bans LLM-driven agentic bots. Lista is exposed on four of the five, and the weekly "marketplace fixes" releases are the running cost.

The real differentiator is the camera work. Automatic garment measurement and ghost-mannequin photography attack returns and buyer questions, a pain the funded cross-listers ignore, and the one thing here no platform gives away free.

Base decomposition

• Market and competition 2.5: a lane with a proven ceiling, where the best-funded competitor now charges nothing and the survivors are all bootstrapped.
• Product and wedge 4.5: the broadest feature set in the batch, measurement and ghost-mannequin genuinely differentiated, all resting on unofficial integrations.
• Traction 2.0: 126 ratings at 4.61 in forty-nine months, no website, no marketing, no Android build, an unsupported "thousands of resellers" claim.
• Unit economics 3.5: no payroll, but generative cost on every item against an undisclosed subscription is the least favourable margin shape here.
• Team 3.0: one skilled Swift engineer, five and a half years on the problem, no co-founder, outside contract work mid-project, self-described as developer not founder.
• Founder trust and disclosure 3.5: identity and history verify cleanly with nothing adverse, offset by no entity, no real website and an unsupported user claim.

Mean = (2.5 + 4.5 + 2.0 + 3.5 + 3.0 + 3.5) / 6 = 19.0 / 6 = 3.17. Adjust +0.25 because four and a half years of unbroken shipping, with a release the week of this review, on zero capital and zero marketing, is durability the rating count understates. Base = 3.42.

Formula

3.42 + 1.00 (EV, Harmonic funding_total $0, $0 to $1M band) + 1.00 (Geo, USA) + 0.00 (Chain) = 5.42.

For Daxos, price and verdict

No investable vehicle. Harmonic records zero rounds, no valuation, no valuation_info, currently_raising null. The App Store seller is an individual, so there is no cap table and no entity to buy into, and no evidence the developer wants capital or a co-founder.

If one were constructed, the check would buy a solo-built cross-lister in the lane that has never produced a venture outcome, against a free product from the best-funded incumbent, on unofficial integrations with four marketplaces that have each broken or banned tools like it in the last two years. It cannot return a fund. DAXOS = 2.0. VERDICT PASS.

What is worth remembering is the camera work, not the company. If single-photo measurement and ghost-mannequin generation ever sit inside a demand network rather than a tool, that is worth something.

Threat to Moe: 1. Overlapping function, 126 ratings, one person, no distribution.

What could not be checked

Prices and tiers, subscribers, revenue, retention, listings processed. Whether any legal entity exists behind the name. Whether an Android build exists. How the integrations are implemented and whether any are official. The review stream directly, so the support reports remain second-hand. Whether the developer now works on Lista full time.

back to the ranked table

Profitable bootstrapped cross-lister with the best distribution in its lane and no identifiable founder anywhere in the public record.

What it is

PrimeLister is a cross-listing and marketplace-automation tool for resellers. One inventory, published to Poshmark, eBay, Mercari, Depop, Etsy, Facebook, Grailed and Shopify, with auto-delist when an item sells, bulk price updates, tags and groups, and scheduled tasks. Alongside the cross-lister it sells separate automation bots for Poshmark, Depop and eBay: auto-share, community share, auto-relist, auto-offer to likers, follow and unfollow. It runs as a Chrome extension plus a web app plus mobile apps (primelister.com, checked 2026-09-16).

Founders and team

This is the problem with the company as an investment, so it goes first. No founder is named anywhere.

Checks run: Harmonic company record and both attached person records; the site, where /about and /pricing both return 404 and no team page appears in the sitemap; the terms of service, which names no officer and only specifies Delaware jurisdiction; the Chrome Web Store publisher page; the App Store listing; and web search on the company and the publishing entity. No adverse record found in the checks run: the Chrome Web Store states "The publisher has a good record with no history of violations", and no litigation, dissolved entity or disputed claim surfaced.

What is visible: the App Store and Chrome Web Store publisher is BYTEBLAST DIGITAL LTD, a name that appears nowhere on the website (iTunes lookup 6478108527 and chromewebstore.google.com, both 2026-09-16). The Harmonic HQ, 251 Little Falls Drive, Wilmington, is a registered-agent address rather than an office. The only two people Harmonic attaches are Harun Seven, a developer in Mersin, Turkey, listed from 2016, and Tuba Canovali, listed as "Stajyer", Turkish for intern, from 2018 (Harmonic persons 129625802 and 179378811, 2026-09-16). Both records start before the 2019 founding date, so confidence is low. The honest summary is a Delaware shell over a small Turkish engineering team, operated anonymously.

Nothing wrong with that for a bootstrapped software business. It is disqualifying for a fund that has to underwrite people.

Product and traction

Chrome extension: 9,000 users, 4.7 stars from 77 ratings (chromewebstore.google.com, 2026-09-16). iOS "Poshmark Bot: PrimeLister": 2,349 ratings at 4.88 average, released 20 February 2024, version 1.0.30, last updated 17 March 2026, which is six months stale (iTunes, 2026-09-16). Also on Google Play. Harmonic web traffic 52,770, the highest of any cross-lister in this batch. Twitter 19,200 followers, LinkedIn 1,762. The homepage claims "Trusted by 10,000+ resellers", consistent with the 9,000 extension installs.

A public roadmap, changelog and feedback board at roadmap.primelister.com plus a docs site is more product hygiene than most tools in this lane show. Against that, an automation tool whose iOS build has not shipped in six months, in a lane where Mercari silently broke cross-listers in August 2024 and Poshmark suspended cross-lister users twice in 2025, carries real maintenance risk.

Business model and unit economics

Cross-listing is $49.99 a month or $399.99 a year. The Poshmark bot is $25 per closet per month, or $20 paid yearly. Seven-day free trial, no card required, Stripe billing (primelister.com/pricing/cross-listing and /llm-resources/pricing.md, both 2026-09-16).

That pricing is aggressive given Flyp, the best-funded cross-lister ever at $15.8M, now gives the product away free. PrimeLister can charge it because it never took money and has a low-cost engineering base, so fixed cost is small and almost any paid conversion off 9,000 installs is profit. Two thousand accounts at a blended $35 would be roughly $840k of annual revenue against a four-person cost base in Turkey. A genuinely cash-generative small business. Not a venture business.

Position in this market

Squarely in the tool lane, which has a hard ceiling and has never produced a venture outcome. The survivors are bootstrapped by design: List Perfectly at $0, Crosslist at $0, PrimeLister at $0, Vendoo at $125k. PrimeLister is arguably the best-distributed of them by web traffic and iOS ratings, which confirms the shape of the market rather than escaping it. Everything it does depends on browser automation against platforms that have no official consumer listing API, actively break automation, and in eBay's case banned LLM-driven agentic bots in the February 2026 user agreement.

Base decomposition

Market and competition 3.0: the tool lane with a proven ceiling, a price war started by a free Flyp, and no venture exit in the category's history. Product and wedge 4.0: eight marketplaces, real inventory sync, three separate automation bots, public roadmap and docs, but all of it browser automation with no official API and no AI listing wedge. Traction 4.5: 9,000 extension users, 2,349 iOS ratings and 52,770 monthly visits make it the best-distributed cross-lister in this batch. Unit economics 5.5: $49.99 a month plus a $25 bot, against a four-person low-cost team and zero capital raised, is very probably profitable. Team 2.5: four people, no identifiable founder, no identifiable technical lead. Founder trust and disclosure 2.5: anonymous operator, publisher name that does not match the brand, registered-agent HQ and no team page; nothing adverse found, but nothing to underwrite either.

Mean = (3.0 + 4.0 + 4.5 + 5.5 + 2.5 + 2.5) / 6 = 22.0 / 6 = 3.67. Adjust minus 0.25 because the six-month-stale iOS build compounds the anonymity: there is no one to ask what the maintenance plan is. Base = 3.42.

Formula

3.42 + 1.00 + 1.00 + 0.00 = 5.42.

The geo bonus applies USA because the recorded HQ and legal jurisdiction are Delaware. The operating locus looks like Turkey, which would carry a smaller bonus, and that ambiguity is flagged rather than silently resolved.

For Daxos, price and verdict

Not raising, and structurally not raiseable in its current form. Harmonic records zero rounds and zero investors, the founders are anonymous, and the business is profitable enough not to need money. There is no entry price to negotiate because there is no counterparty to negotiate with. Even if a price appeared, a $100k to $500k cheque into the best-run cross-lister in a lane whose best-funded participant burned $15.8M and then made the product free cannot return a fund; the realistic outcome is a small cash-flow business or a sub-$20M trade sale to a consolidator. Daxos cannot add distribution, platform relationships or hiring help to a team it cannot name. Verdict PASS.

Worth recording for competitive purposes rather than investment: PrimeLister's 9,000 installs and 2,349 iOS ratings put it well ahead of most tools in this lane, so threat to Moe is 3 rather than the default 2.

What could not be checked

Revenue, paying subscriber count, churn and conversion from the 9,000 free installs are all unknown. Founder and officer identity could not be established. Whether ByteBlast Digital LTD is a UK, Turkish or other entity, and how it relates to the Delaware registration, could not be determined. Google Play install count was not retrieved. Whether the six-month gap in iOS releases reflects a stable product or a stalled one is unresolved.

back to the ranked table

A bootstrapped Australian alerting tool telling car dealers about under-market private-party vehicles the second they list, with the best revenue per user in this batch and a single non-technical owner on top of a scraping stack.

What it is

Real-time marketplace alerting for buyers of private-seller used cars. The site (2026-09-16) says it scans Facebook Marketplace, Craigslist, OfferUp, Kijiji, Gumtree, Nextdoor and eBay and pushes a notification the moment a matching vehicle appears, so the dealer messages first.

Adjacent is the right word. The primitive matches the resale tools, watch a marketplace and price the thing, but the object is a car, the customer is a dealer, and the payoff per correct answer is a few thousand dollars of gross profit rather than eight dollars on a jacket. Threat to Moe is 1.

Founders and team

Rory McLeod, Company Owner since August 2022, described in his Harmonic record as part owner and founder, located in Australia, and the primary executive contact (flip@getswoopa.com). His background is film: Swinburne University of Technology, Film and Television; freelance cinematographer since February 2015 and still current; director and cinematographer on a 55-minute SBS documentary, When the River Runs Dry (2018-2019); colourist on History Bites Back for SBS (2021). No engineering degree, no shipped-code history, no GitHub surfaced. On the record, non-technical.

Harmonic lists exactly one person, so no co-founder, CTO or lead engineer is identifiable. His own phrase "part owner" implies another owner named nowhere in the sources used.

Checks run: Harmonic company record and people list, Harmonic person record for McLeod with dated history and education, the site including pricing, the iTunes search and lookup endpoints for all three apps, and Harmonic social links. No adverse record found in the checks run. LinkedIn scraping was not run; Harmonic carried the full history.

Product and traction

Three iOS apps, all Market Sourcing Solutions PTY LTD, checked 2026-09-16. Swoopa: released 2024-01-31, updated 2026-09-06, version 1.30.6, 541 ratings at 4.40. Swoopa Dealers: released 2025-01-23, updated 2026-09-06, 53 ratings at 4.32. Swoopa Reposter: released 2025-09-01, last updated 2025-09-09, 2 ratings at 1.00, abandoned after one patch.

Two and a half years of continuous shipping into version 1.30 is the strongest product signal in this batch. Five hundred and forty-one ratings is small in consumer terms but large for a B2B tool at three figures a month. The site carries eleven languages.

Business model and unit economics

Pricing is published in full (getswoopa.com/pricing, 2026-09-16), which almost nothing else in this scan does. Committed: Go $28 a month on 12 months at $336, Turbo $99 on 6 months at $594, Nitro $199 on 6 months at $1,194. Rolling monthly: $47, $144, $352. Seven-day trial. Tiers differ on alert latency, marketplace coverage (Go is Facebook Marketplace only) and AI VIN and title detection.

Blended ARPU near $100 a month against six people means a few hundred paying dealers fund the business, consistent with 541 ratings and no capital in four years.

The unhedged risk is the input. Facebook Marketplace is the primary source and has never exposed a public C2C listing API, so this is scraping. The brief documents what follows when a platform closes that door: Mercari's August 2024 change silently broke the cross-listers, Poshmark suspended cross-lister users twice in 2025, eBay's February 2026 user agreement bans LLM-driven agentic bots. The most expensive tier is sold on instant alert speed, the thing that breaks first. None of this is disclosed.

Position in this market

The healthiest revenue model in this batch, in a different market. It confirms the brief's economics from outside: the tool lane works in front of a buyer with real margin and fails in front of a consumer pricing a thrifted jacket, and it shares the lane's weakness, that the product is downstream of platforms that owe it nothing.

Base decomposition

Market and competition 5.0: large, high-value, underserved B2B niche where the buyer has budget, though ACV, CarOffer, Marketcheck and a tail of Facebook scrapers crowd it. Product and wedge 5.0: real depth across alerting, plate-to-VIN, comps, history and CRM on an actively shipped 1.30 codebase; the wedge is latency, rented not owned. Traction 4.5: 541 and 53 ratings, four years alive, updates through September 2026, and one abandoned app at 1.0 stars. Unit economics 5.5: $28 to $352 a month against six people, bootstrapped four years, payback for the customer on a single vehicle. Team 2.5: one identifiable person, a working cinematographer with no engineering record, no named co-founder, no technical leadership visible. Founder trust and disclosure 4.0: full public pricing under a named legal entity, but no about page, no team, and no acknowledgement that the data supply depends on scraping.

Mean of 5.0, 5.0, 4.5, 5.5, 2.5, 4.0 is 26.5 over 6 equals 4.42. Adjusted down 0.25 for the undisclosed scraping dependency, the risk most likely to end the company and the one the pricing page monetises hardest. BASE 4.17.

The formula line

4.17 + 1.00 (EV, $0 raised) + 0.25 (Geo, Australia) + 0.00 (Chain) = 5.42

For Daxos, price and verdict

PASS. Daxos fit 2.5 out of 10. No evidence of a raise: $0 across four years, no investors, no round. Australian, founder running it alongside a freelance film career, no technical leadership to underwrite, and the wrong sector for the thesis this scan tests.

The narrow case for a call, short of MEET: Swoopa has solved the pricing problem the consumer lane cannot, and how it holds up alerting against Facebook Marketplace at scale is useful intelligence. That is research, not diligence.

What could not be checked

Revenue, paying-customer count, churn and tier mix are undisclosed, so ARR is an inference from ratings and price. Harmonic returns no location, so the Australian HQ rests on the App Store publisher entity and the founder's location, not a registry. The other owner implied by "part owner" could not be identified. No engineering lead is identifiable, so whether the scraping stack is built or bought is unknown.

back to the ranked table

A well-built authenticated watch marketplace with a 4.89-star app and 33,764 live listings, whose last money was a $670k bridge labelled a Series A.

What it is

Bezel is an authenticated marketplace for pre-owned luxury watches. Every watch is authenticated in house before it ships, with diagnostic testing, insured overnight shipping, buyer protection, a concierge team, auctions, free appraisals and a "Collect" surface for tracking a collection (getbezel.com, read 2026-09-16). Scope check passed: $8,670,000 total, under the $10.9M threshold, privately held, no acquirer.

Founders and team

Quaid Walker, co-founder and CEO. Design lead and founding designer on Google TV before leaving to start Bezel after a bad first Rolex purchase (wornandwound.com and karagoldin.com interviews, read 2026-09-16). A product designer rather than an engineer, which shows in the app. Darryl Johnson, co-founder and CTO. Chase Pion, co-founder and COO (tracxn.com profile, read 2026-09-16). A three-way split of product, engineering and operations is right for an authenticated marketplace, where operations is most of the work.

Five years on the same problem with the same team, headcount now 34. The investor list is strong for the amount raised: BoxGroup, Michael Rubin, Courtside Ventures, Operator Partners, Shrug Capital and Abstract.

Checks run: Harmonic enrich, company site, iTunes search and lookup, web search on all three founders, Tracxn and CB Insights profiles, podcast and trade press interviews. No adverse record found in the checks run: no litigation, no dissolved entities, no disputed claims, no inflated titles.

Product and traction

The strongest live product in this batch by a distance. 33,764 luxury watches listed as of 2026-09-16. The iOS app has 3,020 ratings at 4.89, released 2022-03-08 and updated 2026-09-15, so it is actively shipped. Forty-plus brands from Rolex down to MB&F, which means real dealer supply.

The comparison that matters is the incumbent. Chrono24 has 68,821 iOS ratings. Bezel has 3,020, four percent of that, after four and a half years; Hodinkee 9,261, The 1916 Company 6,285. Bezel's average beats all of them, which says the product is better; the count says it has not won distribution.

Business model and unit economics

Marketplace take rate on high order value transactions, undisclosed, plus concierge and auctions. High AOV is the friendly part: a five to ten percent take on a $10,000 watch is real revenue per order. The unfriendly part is that authentication, diagnostic testing, insured shipping and human concierge scale close to linearly with volume. Thirty-four people on $8.67M raised means either the business generates meaningful revenue or the raise is largely spent, and no public figure tells which. The market is a headwind too: secondary watch prices have fallen from the 2022 peak, and WatchCharts' indexes have tracked that decline for years.

Position in this market

Bezel is a demand network, the right side of the dividing line, in a vertical with liquid comps. But it is the most contested and most capitalised vertical in resale: Chrono24, Watchfinder inside Richemont, Hodinkee inside Watches of Switzerland, The 1916 Company, Bob's Watches, and eBay's Authenticity Guarantee, which made authenticated watch selling free on the largest demand pool on earth. Threat to Moe: 1.

Base decomposition

• Market and competition: 3.5. Large GMV market, but the most crowded and best-funded vertical in this set, with corporate-owned incumbents and a multi-year price decline.
• Product and wedge: 5.0. Authentication plus concierge is the correct trust wedge, executed well; 4.89 stars and a same-week update are earned.
• Traction: 4.5. 33,764 live listings is substantial supply. 3,020 iOS ratings in four and a half years against Chrono24's 68,821 is modest demand.
• Unit economics: 3.5. High AOV helps, but authentication, shipping and concierge are linear labour costs, and 34 heads on $8.67M is tight.
• Team: 5.5. Ex-Google TV design lead plus a real CTO and COO, five years in, well backed.
• Founder trust: 4.0. Nothing adverse, but no GMV or revenue is public and the last round's framing is generous.

Mean = (3.5 + 5.0 + 4.5 + 3.5 + 5.5 + 4.0) / 6 = 26.0 / 6 = 4.33. Adjust minus 0.25: a $670,000 last round total labelled SERIES_A in May 2025 while carrying 34 employees reads as a bridge, and a bridge after $8M is a signal about the A that did not clear. Base = 4.08.

The formula

4.08 + 0.25 (EV, $8.67M, the $5M to $10M tier) + 1.00 (Geo, USA) + 0.00 (Chain) = 5.33.

For Daxos, price and verdict

Harmonic estimates a $24M post, marked ESTIMATED, derived from the $670k round and adjusted by prior rounds, so the number is soft and probably understates the last priced mark. At $24M or above, a $500k check buys around two percent of a company that has to take share from Chrono24 in a declining market, with a labour-heavy cost structure and a bridge behind it. Daxos has no watch supply, no dealer relationships and no authentication expertise.

WATCH rather than PASS, because the reasons to decline are price and category rather than quality, and the re-open condition is concrete: a genuinely priced Series A with disclosed GMV, take rate and gross margin, at a valuation reflecting the market decline rather than the 2022 peak. Absent it, no check. DAXOS fit 3.0.

What could not be checked

GMV, revenue, take rate, gross margin, burn and runway. Whether the May 2025 $670k was a priced round, a SAFE or an extension. Google Play presence. Cap table and preference stack. No LinkedIn scrape: the Apify actor errored twice and was stopped.

back to the ranked table

By far the largest consumer audience in this batch, roughly 16,900 iOS ratings at 4.82 and a claimed 750,000 users for antique identification, run by an anonymous LLC that stops at the answer and never touches the transaction.

What it is

Curio is a free antique and vintage identifier. Photograph a piece and it returns the maker, era, country of origin, the history of the object, an estimated value range drawn from online listings, and visual matches you can check yourself (curio.app, fetched 2026-09-16). The site pitches three users: thrifters deciding at the register, homeowners and inheritors cataloguing a house, and dealers and auctioneers pricing a booth or an estate.

It identifies and values. It does not list, cross-list or transact. That makes it adjacent to the cross-listing lane, and it puts it in the vertical-with-comps corner of the market rather than the commoditized general-purpose corner.

Founders and team

No founder is named. The App Store seller is Dionysus Labs, LLC. curio.app has no about page (/about returns 404), no team, no address and no contact email in the page source. The footer reads "© 2026 Dionysus Labs, LLC". dionysuslabs.com is a parked domain-squatter page, confirmed through Harmonic, which returns the generic "your first and best source for all of the information you're looking for" text.

Checks run: Harmonic enrich on curio.app and on dionysuslabs.com, the App Store developer record for artist id 1738854903, the site root plus /about, /pricing, /terms and /privacy, and a web search on the entity name. No named human surfaced. No adverse record was found, and again that is because there is no name to search.

Two things are worth recording as disclosure quality rather than as fraud. The /terms page serves Apple's standard Licensed Application End User License Agreement instead of the company's own terms. The privacy policy is an unedited generator template, including the phrase "as a a Freemium service". Neither is illegal; both say this is an operator that has not built a corporate face.

The same developer shipped TreeID (235 ratings) and Animal Identifier (29 ratings) within six weeks of Curio in 2024. Both are abandoned, TreeID since 2025-05 and Animal Identifier since 2024-11. That is the identifier-app-mill pattern. The difference from the others in this batch is that this operator picked the winner and has shipped it for 26 months straight, now at v1.10.0 dated 2026-09-10.

Product and traction

About 16,900 iOS ratings across five storefronts, averaging 4.75 to 4.82, from a 2024-07-22 release. The site claims "750,000+ users", which is unverifiable but directionally consistent with that rating count at a normal one-to-two percent rating rate. Google Play package is com.dionysus_labs.antiqueid.

The competitive number matters more than the absolute one. AntiqSnap, by Next Vision Limited, released 2025-10-15 and already has 31,338 US ratings at 4.68 as of 2026-09-16. It reached double Curio's US rating count in 11 months against Curio's 26. Whatever Curio built, a fast follower copied it and is out-distributing it.

Business model and unit economics

Freemium subscription. The site's call to action is "Identify my first item free", so the paywall sits immediately after the first scan. The privacy policy confirms a freemium service. No price is published anywhere on the site.

The structural weakness is repeat use. You identify a grandmother's clock once. Unlike a reseller tool with weekly listing volume, an identify-only consumer app has to keep buying new installs forever, which makes it a paid-acquisition treadmill rather than a compounding account base. Curio sits on the exact moment a user learns an object is worth $1,240 and captures none of the transaction that follows.

Position in this market

Antiques is the best consumer vertical in this batch. The comps are illiquid, the research genuinely costs twenty minutes a piece, and the user cannot guess the answer, which is the opposite of a used iPhone. The category brief's observation holds here: where this market compounded, it was vertical with comps. But the companies that compounded, CollX and Alt, went on to own the transaction or the asset. Curio stopped at the answer and monetizes it with a consumer subscription, which is why it has 750,000 users and no obvious path to a venture outcome.

Base decomposition

• Market and competition 3.0. Real and proven consumer demand in a vertical with genuine expertise asymmetry, but AntiqSnap doubled Curio's rating count in 11 months, and Google Lens plus free ChatGPT do a passable version.
• Product and wedge 4.0. The right vertical and a well-executed answer, undermined by stopping before the transaction.
• Traction 6.5. About 16,900 ratings at 4.82 across five countries, 750,000 claimed users and 26 months of continuous shipping. The strongest consumer result in the batch by a wide margin.
• Unit economics 4.0. Freemium subscription that works, on a use case with almost no natural repeat, so it is a paid-acquisition treadmill with no transaction capture.
• Team 1.5. Anonymous LLC with two abandoned sibling apps, credited for two years of sustained execution on the one that worked.
• Founder trust and disclosure 1.5. No named human, no about page, Apple boilerplate served as the company's terms, a template privacy policy with an uncorrected typo, and a parked corporate domain.

Mean = (3.0 + 4.0 + 6.5 + 4.0 + 1.5 + 1.5) / 6 = 20.5 / 6 = 3.42. Adjusted -0.17 because the AntiqSnap comparison shows the position eroding against a faster clone in real time. Base = 3.25.

Formula

3.25 Base + 1.00 EV ($0 raised) + 1.00 Geo (USA, inferred from the LLC form and US storefront concentration) + 0.00 Chain = COMPANY 5.25.

For Daxos

Pass, but for a different reason than the rest of the batch. This is the one company here with a genuinely valuable asset: 750,000 consumers who have handed it a photo at the exact moment they are deciding whether to sell something. In the hands of an operator who wanted to own the transaction, that is the top of a real funnel.

It is still uninvestable. There is no named founder, no about page, no corporate domain, no LinkedIn and no round, so there is no counterparty to diligence and no one to have a conversation with. That also rules out MEET, which requires a founder to meet. If the operator ever surfaced with a name and a raise, this would flip to MEET immediately on the strength of the funnel. Until then, DAXOS 2.5, VERDICT PASS.

What could not be checked

Revenue, subscription price, MAU, retention, churn, install spend, and the state of registration for Dionysus Labs, LLC. Harmonic's curio.app record is a stale mismatch, describing a one-person Thousand Oaks search product founded in 2016 with the tagline "Bring the ultimate search to all your content", which is not this business, so HQ, founding date and headcount are all unknown rather than merely unreported. The 750,000 user figure is the company's own claim and was not independently verified.

back to the ranked table

A 23-year-old Toronto multichannel listing business serving eBay Motors auto-parts sellers, never raised, 15 staff, marketing frozen since 2022, and structurally irrelevant to consumer resale.

What it is

Multichannel listing, inventory, order and warehouse software for professional sellers. The homepage (2026-09-16) calls it the complete eCommerce solution for the day-to-day of eBusiness, integrating eBay, eBay Motors, Amazon, Walmart, Newegg, Etsy, Reverb, Shopify and a hosted Kyozou Webstore.

The vertical it actually lives in is auto parts. eBay Motors is a first-class integration, and the only case study on the resources page is a Nissan dealership parts department that grew sales 20 percent.

Harmonic (2026-09-16): B2B, Logistics and Supply Chain Technology, founded 2003, Toronto, headcount 15, funding_total $0.00 across 0 rounds, no investors, VENTURE_UNKNOWN. Twenty-three years old and never raised.

Founders and team

Larry Velman, founder. Harmonic's person record is nearly empty: one entry, Founder at Kyozou with no dates, education at Seneca Polytechnic 2000-2003, LinkedIn linkedin.com/in/larry-velman-4b404b5. For a founder of 23 years that thinness is the finding.

Alexei Smirnov, Harmonic-listed as CTO, BD, Co-Owner and Founder from October 2003 to October 2008, the technical origin of the product. His career since is more prominent than the company he left: CTO at Dealitlive (2008-2011), Director of Technology at Cellfire, product owner and senior architect at eBay (2010-2012), CEO of SwapMyGames (2009-2016), ten years at ReturnPro to VP of Engineering (2012-2022), then DevOps and AI roles at Doxel, Ford and VirtuMAN. Gone 18 years.

Reid Armstrong, General Manager since November 2017 after four years inside the company. Harmonic also shows him holding Director of Operations at CureDiva from January 2018 to now, concurrent with the GM role. On the record the general manager has held a second job for eight years.

No current CTO or lead engineer is identifiable.

Checks run: Harmonic company record and people list, Harmonic person records for Velman, Smirnov and Armstrong with dated histories and education, the site including resources and sitemap, and the iTunes search endpoint. No adverse record found in the checks run. LinkedIn scraping was not run because Harmonic returned complete histories.

Product and traction

Pricing is stated once, on the homepage: starting at $199.99 a month based on a percentage of sales, no contracts, quote by demo (2026-09-16). There is no pricing page; /pricing/ returns 404.

Maintenance signals are mixed. The WordPress sitemap index (2026-09-16) shows post-sitemap last modified 2022-08-16, portfolio items 2019-03-20, integrations 2024-04-06, pages 2026-07-27. Fifteen employees and a live login mean the product is operating and supported: a mature business coasting, not a dead one. Status live.

No iOS or Google Play app, so the rating-count proxy does not exist here.

Business model and unit economics

Monthly SaaS at a $199.99 floor plus a percentage of customer sales. Fifteen staff, 23 years, zero outside capital means it has covered payroll for two decades. Growth evidence is absent: no published customer count, GMV or ARR, and a marketing engine that stopped in 2022.

Position in this market

Sits with Listing Mirror and Sellbrite in the professional lane, the most specialised and oldest of the three. The brief's rule holds: every survivor in tools is bootstrapped (List Perfectly, Crosslist, PrimeLister, Listing Mirror, Kyozou all $0), and the funded ones are the graveyard. Threat to Moe is 1.

Base decomposition

Market and competition 3.0: real but narrow and structurally flat, and the auto-parts-on-eBay-Motors niche has lost share for a decade. Product and wedge 4.0: genuine operational depth and 23 years of integration work, but the technology is old and the integration set has not moved since April 2024. Traction 4.5: 15 staff and two decades of survival is a real business, though there is not one third-party-verifiable metric anywhere. Unit economics 4.5: percentage-of-sales on a $199.99 floor is a sound model and self-funding for 23 years proves the margin, but there is no evidence of growth. Team 2.5: the technical co-founder left in 2008, no current engineering leadership is identifiable, the founder has no visible record beyond this company, and the GM has held a concurrent role elsewhere since 2018. Founder trust and disclosure 4.0: nothing misleading, but no team page, no pricing page, no metrics, and an effectively invisible founder.

Mean of 3.0, 4.0, 4.5, 4.5, 2.5, 4.0 is 22.5 over 6 equals 3.75. Adjusted down 0.25 for a 23-year-old codebase, no identifiable technical leadership and marketing that stopped in 2022. BASE 3.50.

The formula line

3.50 + 1.00 (EV, $0 raised) + 0.75 (Geo, Canada) + 0.00 (Chain) = 5.25

For Daxos, price and verdict

PASS. Daxos fit 0.5 out of 10. There is nothing to buy. A 23-year-old bootstrapped company with 15 staff is not raising a seed round, and no price would work: $100k to $500k into a flat niche SaaS cannot return a seed fund, and the lane's only realised exit was Sellbrite into GoDaddy at an undisclosed price on $2.9M raised. The lasting value of the record is as evidence for the brief: a quarter century in the tool lane produced a stable 15-person company and no venture outcome.

What could not be checked

Revenue, ARR, customer count and churn are undisclosed, and the percentage rate above the $199.99 floor is quote-only, so the effective take rate is unknown. Google Play was not checked because the site surfaces no mobile product.

back to the ranked table

Bootstrapped Poshmark bot and cross-lister run by a husband-and-wife team across six brands, shipping actively, and wholly dependent on automation Poshmark can switch off.

What it is

Posh Sidekick, sold in the App Store as Sidekick Tools, automates a Poshmark seller's daily grind: self-shares, party shares, community shares, follow and unfollow to grow a closet, bulk and automatic offers to likers, daily relisting, and daily and weekly statistics. On top of that sits a cross-posting product covering Poshmark, eBay, Mercari, Depop, Etsy, Grailed and Whatnot, with auto-delist when an item sells, bulk editing, and on higher tiers AI stock photo creation, image background removal and sales reporting (poshsidekick.com/pricing, retrieved 2026-09-16; App Store listing 1589949924, 2026-09-16).

The homepage itself returned HTTP 429 rate limiting on direct fetch, which is a small sign of how aggressively the site throttles automated traffic.

Founders and team

Michael Chierchio is CEO and co-founder since November 2020, based in Brooklyn. Erica Chierchio is co-founder, listed from January 2018, NYU business and marketing background (Harmonic persons 171392487 and 171392638, 2026-09-16). The company description says the pair have been Poshmark resellers since 2012 and the App Store copy says "We've been doing this since 2015". Fourteen years in the community and eleven building for it is unusual persistence and the strongest thing on this card.

Neither shows an engineering degree, shipped-code history or GitHub presence in any check run. The iOS app is published under "Michael Chierchio" as a personal seller account, not a company one.

Both run several things at once. Michael's record lists Lyn Media Group from 2017, Crosslisting.com as CEO and co-founder from November 2020, and Reseller Vault, Mercaribot, Poshmark Pro Tools and Lyn Financial, a credit repair service, all from 2022. Erica's lists co-founder of Crosslisting.com from March 2022 and CEO of ZORS from April 2020. Posh Sidekick is one brand inside a portfolio of reseller-tool properties, and founder attention is split across at least six entities.

Checks run: Harmonic company and both person records, the site and pricing page, the App Store listing, LinkedIn company pages for Posh Sidekick and Poshmark Pro Tools, ZoomInfo and SignalHire profiles, and targeted searches for litigation between Poshmark and either the founders or the tools. No adverse record found in the checks run: no lawsuit, cease-and-desist, dissolved entity or disputed claim surfaced. The credit repair venture is noted only because that sector draws regulatory attention; nothing specific was found against it.

Product and traction

iOS app 2,268 ratings at 4.66 average. Released 6 December 2021, version 2.8.9 shipped 15 September 2026, the day before this check (iTunes lookup 1589949924, 2026-09-16). That is an actively maintained product, which in this lane matters more than it sounds: tools that stop shipping die the next time a platform changes its markup. Harmonic web traffic 31,180. Instagram 8,438 followers, LinkedIn 215, Twitter 18. The pricing page claims "trusted by 10,000+ resellers".

Four and a half years to 2,268 ratings is modest but real, and puts Sidekick in roughly the same band as PrimeLister at 2,349 and Vendoo at 2,784, below ClosetMate at 7,337 (iTunes, 2026-09-16).

Business model and unit economics

Two ladders, both capped by volume. Cross posting runs $9.99 a month for 100 items, $19.99 for 250, $29.99 for 500 and $44.99 for 1,000. Automation runs $29.99 for 50 crosslists, $49.99 for 150 and $59.99 for 250 (poshsidekick.com/pricing, 2026-09-16). Item-count tiers are sound design because revenue expands as a closet grows rather than sitting flat like a seat price.

The cost base is four people with no capital raised, so margin is presumably high and the business presumably profitable. The problem is the foundation, not the margin. Every dollar comes from automating a platform that does not permit automation. Poshmark suspended cross-lister users in May and again in November 2025, and there is no official consumer listing API to fall back on. One enforcement decision removes the product.

Position in this market

This is the most platform-exposed corner of the tool lane, which itself has the proven ceiling. The bootstrapped survivors, List Perfectly at zero raised, Crosslist at zero, PrimeLister at zero, Vendoo at $125k, all look like this, and Posh Sidekick belongs with them. It is priced into a market where Flyp, having burned $15.8M, now gives the same core function away free, which caps what any of these tools can charge. It has none of the things that have produced outcomes here: no demand network, no supply, no checkout, no liquid-comp vertical underneath.

Base decomposition

Market and competition 2.5: tool lane, hard ceiling, free competitor at the top of the funding table, and the most enforcement-exposed subsegment of it. Product and wedge 4.0: seven marketplaces, genuine automation depth, AI photo tools on the higher tiers, and a build shipped the day before this check. Traction 4.0: 2,268 iOS ratings, 31,180 monthly visits and 8,438 Instagram followers, competitive with the other bootstrapped tools. Unit economics 4.0: volume-tiered pricing with natural expansion on a tiny cost base, undercut by total dependence on prohibited automation. Team 3.0: fourteen years in the community and eleven building for it, but no engineering credential and attention split across six ventures. Founder trust and disclosure 3.5: founders named and publicly identifiable with an about-us page and a long track record, nothing adverse found, but one product sold under three brand names and published from a personal developer account.

Mean = (2.5 + 4.0 + 4.0 + 4.0 + 3.0 + 3.5) / 6 = 21.0 / 6 = 3.50. Adjust minus 0.25 because the flagship product is explicitly a Poshmark bot in a year when Poshmark has already suspended cross-lister users twice, and that single-point failure is not reflected in any of the six sub-scores strongly enough. Base = 3.25.

Formula

3.25 + 1.00 + 1.00 + 0.00 = 5.25.

For Daxos, price and verdict

There is nothing to buy. Harmonic records zero rounds and zero investors, and a profitable husband-and-wife operation running six brands has no reason to sell equity or take a board member. If a price did appear, the underwriting question would be whether a fund should pay for a revenue stream that a single Poshmark enforcement decision can zero out, in a lane that has never produced a venture outcome, with no technical founder to rebuild on an API basis if automation is closed off. The answer is no at any price a founder would accept. Daxos brings no platform relationships, no reseller distribution and no engineering help. Verdict PASS.

For competitive tracking rather than investment: 2,268 iOS ratings, seven marketplaces and shipped AI photo tooling make this a more real competitor to Moe than the batch default suggests, so threat is 3.

What could not be checked

Revenue, paying subscriber count, churn and the split between the cross-posting and automation ladders are all unknown. The homepage and pricing page could only be read through a proxy fetch because the site returned HTTP 429 to direct requests, so some site content was not seen. The legal entity behind the brands, and how Lyn Media Group, Crosslisting.com, Poshmark Pro Tools, Mercaribot and Reseller Vault relate to each other in ownership terms, could not be established. Google Play install count was not retrieved. Whether Poshmark has ever sent enforcement correspondence to these founders is unknown; nothing public was found.

back to the ranked table

Spanish instant-cash buyer of used books, games and media with free home pickup and real consumer traffic, but a thin-margin inventory business on 4.7M across six rounds in six years.

What it is

Hamelyn is not a listing tool. It buys. A user scans a barcode on a used book, game, CD or film, gets an instant price for the lot, Hamelyn collects the boxes from the home for free, and pays into the bank account within 24 hours (hamelyn.com, fetched 2026-09-16). Hamelyn then resells on its own storefront and on the main marketplaces. This is the momox, Ziffit and Decluttr model applied to Spain. It is adjacent to Moe in that both answer "what is this worth", but Hamelyn answers with a binding offer and takes the goods, which is a different business with a warehouse in it.

Founders and team

Federico Aguilar Calvo, CEO, start 2020-07. Harmonic: Senior Data Scientist at Servihabitat 2019 to 2020, Data Science Consultant at Minsait 2017 to 2019, research at UPC, Data Analyst at Volkswagen Group Retail Spain 2016 to 2017 and at the Andalusian Institute of Technology 2014 to 2015. Education is four Udacity nanodegrees (full stack, deep learning, data analysis, machine learning) rather than a university engineering degree. Technical by practice, in data and pricing rather than product engineering, which is the correct skill for an instant-offer pricing engine. One loose end: Harmonic still shows him as Director de proyectos Data Science at CONSULTANIA ASP SL from 2020-01 with no end date, running concurrently with the CEO role. Either a stale profile entry or divided attention; it could not be resolved.

Valentin Izquierdo, Co-Founder, start 2021-01. Harmonic: Founder of Hooked Mounts 2019-11 to 2020-06 (seven months, no visible outcome), Export Area Manager UK and Italy at Doccia Group 2015 to 2020. BBA from Florida Atlantic University and ESIC Business School. Commercial, not technical.

Neither founder has a prior exit. Both have been on this problem since 2020 and 2021, six years of persistence. The bench is operational rather than product-led: a COO, a Director of Logistics and Warehouse, a Director of Automation, a CMO, quality control staff, one full stack developer in the Harmonic people list. Correct shape for a logistics business, thin for a technology one.

Checks run: Harmonic company and person records, company site, App Store ES listing, general search. No adverse record found in the checks run.

Product and traction

Harmonic web_traffic 218,100. By a wide margin the largest traffic figure in this comparison set, and real consumer demand, driven by Spanish-language SEO on "vender libros usados" which the site is heavily optimised for.

iOS app "Hamelyn" (seller Hamelyn Technologies SL, ES store), 214 ratings at 3.75, released 2024-02-07, updated 2026-09-01 (iTunes, 2026-09-16). Two and a half years to 214 ratings is slow, and 3.75 is weak, typical of instant-offer businesses where users are disappointed by the quoted price. The nearest direct comparator in Spain, Reciclibros (RecycLivre), has 358 ratings at 4.49 since 2022-11. Hamelyn is behind on both count and score against the company doing the same thing in the same country.

Categories have widened from books and films to video games, CDs and vinyl, which is the right move for basket size. Google Play is advertised on the site.

Business model and unit economics

Principal buying. Hamelyn owns the inventory, so gross margin is the spread between the instant offer and resale, less free pickup logistics, grading, storage and the marketplace fee on resale. Free home collection is a genuine consumer advantage over mail-in competitors and also the largest cost per transaction. Every euro of growth consumes working capital.

$4,693,453 across six rounds since 2020, with a July 2025 Series A whose amount Harmonic records as undisclosed, is not enough capital to run an inventory business at scale. Compare momox, which reached a few hundred million euros of revenue in Germany on far more capital. Either Hamelyn is running tight and near breakeven, which would be creditable, or it is capital-constrained and cannot buy. Nothing public resolves which.

Position in this market

Hamelyn is on the correct side of the category brief in one respect: it is a demand network in miniature, not a seller tool, and it owns the transaction rather than renting access to someone else's marketplace. Platform dependence is low because it sells through its own storefront as well as marketplaces. But the ceiling is set by Spain, by used media, and by the margin structure. Vinted and Wallapop own Spanish C2C attention, and free scan-and-price now shipping inside them narrows Hamelyn's reason to exist to the free pickup and the instant cash, which are logistics advantages, not technology ones.

Base decomposition

• Market and competition 4.5: real category, small country, a direct Spanish competitor with more ratings and a better score, and the C2C giants adjacent.
• Product and wedge 5.0: scan, instant price, free pickup, paid in 24 hours is a clean promise and free pickup is a genuine edge over mail-in.
• Traction 5.5: 218,100 monthly visits and six years of operation is the strongest live consumer flow in this batch; 214 ratings at 3.75 undercuts it.
• Unit economics 3.5: thin-spread principal buying with per-lot collection cost and a permanent working capital drain, on very little capital.
• Team 4.5: a data-science CEO who fits the pricing problem, a commercial co-founder with one short prior venture, no exits, six years on the problem.
• Founder trust and disclosure 4.5: no metrics published, Series A amount undisclosed, one unresolved concurrent role on the CEO record.

Mean = (4.5 + 5.0 + 5.5 + 3.5 + 4.5 + 4.5) / 6 = 27.5 / 6 = 4.58. Adjust -0.25: six rounds and six years to $4.7M with an undisclosed Series A in an inventory business reads as grinding rather than compounding. Base = 4.33.

Formula

4.33 + 0.50 (EV, $4.69M in the $3M to $5M band) + 0.25 (Geo, EU) + 0.00 (Chain) = 5.08.

For Daxos, price and verdict

Harmonic shows currently_raising null, no valuation field and no valuation_info at all, so there is no price to anchor on. A Series A that goes unreported usually means it was small or a structured extension. A $100k to $500k check into a Spanish working-capital-hungry used-media buyer cannot return a fund: the realistic best case is a trade sale to momox, Ziffit or a European recommerce roll-up at a revenue multiple on a low-margin business. Daxos brings no Spanish logistics, no inventory finance and no distribution. DAXOS = 2.5. VERDICT PASS.

Threat to Moe: 1. Different country, different category, different mechanism.

What could not be checked

The Series A amount, revenue, gross margin, inventory turns, and whether the business is near breakeven. Google Play install and rating counts. Whether the concurrent CONSULTANIA ASP SL role for the CEO is live or a stale entry. LinkedIn was not scraped; founder history comes from Harmonic person records.

back to the ranked table

A fifteen-year-old paid home inventory app run by a strong security engineer, still shipping updates while its own website serves an untrusted certificate.

What it is

Nest Egg is a suite of home and small-business inventory apps. Scan a barcode, the app looks up the product and fills in the details, take photos, organise. Three iOS SKUs: Inventory at $6.99 one-time, Lite, and In Cloud, in 32 languages. The App Store description claims "More than 7 million items scanned" and the site adds "303 Software Releases, 142 Countries, 4 Thousand Support Tickets" (App Store listing and nestegg.cloud/home-inventory/, 2026-09-16).

It matters here because it is the identification leg of the scan-value-list loop, built in isolation and fifteen years early. It scans and identifies. It never added valuation or listing.

Founders and team

Satya D., Chief Executive Officer (Harmonic person 195265, LinkedIn satya-das2, Bay Area). Harmonic records his tenure as 2017-01 to present but tags the role_type FALSE_POSITION, meaning its own system doubts the association. His person record independently lists the same role, so I treat it as real and flag it.

His career is deep infrastructure security engineering: Senior Software Engineer at McAfee 1999-2002, PGP 2002-2004, Check Point 2004-2005, Tech Lead at Webroot 2006-2011, Architect at Vidder 2012-2013, Director of Engineering at Attivo Networks 2014-2017, later acquired by SentinelOne. A credible technical operator, with no consumer, commerce or marketplace experience anywhere in the record. Headcount is 5, and Harmonic lists no other named people, no co-founder and no investor for the 2021 angel round.

Checks run: Harmonic company and person records, App Store search and lookup for all three SKUs, direct fetch and TLS diagnosis of nestegg.cloud, DNS lookup, web search. No adverse record found. Harmonic's full name for the CEO resolves only to "Satya D.", so founder identification here is less complete than elsewhere in this batch.

Product and traction

US App Store, 2026-09-16. Nest Egg - Inventory: 617 ratings at 4.43, released 2011-05-05, version 4.2.41, updated 2026-05-15, price $6.99. Lite: 170 ratings at 4.49, released 2013-05-10, updated 2026-05-15. In Cloud: 48 ratings at 4.31, released 2018-12-08, updated 2026-08-16. Harmonic web traffic 4,900 monthly, LinkedIn 154 followers, no other social presence recorded.

835 ratings across three apps over fifteen years. The apps are genuinely maintained.

The operational finding is the website. nestegg.cloud resolves to 157.230.72.50 and completes a TCP connection, but the TLS handshake fails validation with "unable to get local issuer certificate" (curl -v, 2026-09-16), so every mainstream browser shows a full-page security interstitial. Content is still served behind it and the App Store still points users there. A company that ships an app update in August and lets its own front door serve an untrusted certificate is not being run as a business. The copy also advertises being "Super Retina HD ready", a phrase from the 2017 iPhone X launch.

Business model and unit economics

A $6.99 one-time paid download. That is the whole problem: a one-time purchase generates no recurring revenue, so the business cannot compound however many people buy it. 617 lifetime ratings implies an install base in the low tens of thousands, which at $6.99 is not a company. The In Cloud SKU launched in 2018 is the attempted fix, and after eight years it has 48 ratings.

Position in this market

Nest Egg occupies the home-inventory wedge alongside Itemtopia and Everspruce, the one Sortly exited in favour of contractors. It demonstrates the brief's central point: barcode scan plus automatic product identification, the hardest-looking part of the consumer loop, is free inside Google Lens, Apple Visual Intelligence and every retailer app. Nest Egg spent fifteen years building it as the product.

Base decomposition

• Market and competition: 3.5. A wedge with a fifteen-year record of not converting, now commoditized by free OS-level visual search.
• Product and wedge: 4.5. Real depth for five people: three SKUs, 32 languages, barcode auto-fill, still shipping in 2026.
• Traction: 2.5. 835 ratings over fifteen years, 4,900 monthly visits, 154 LinkedIn followers. The "7 million scans" figure is lifetime cumulative.
• Unit economics: 2.5. A $6.99 one-time purchase cannot compound, and the subscription SKU has 48 ratings after eight years.
• Team: 4.5. Strong security engineering career. No consumer or commerce experience, no named co-founder, five people.
• Founder trust and disclosure: 2.5. The website fails TLS validation while the App Store sends users to it, the copy is nine years stale, and Harmonic doubts the only person association on file.

Mean = (3.5 + 4.5 + 2.5 + 2.5 + 4.5 + 2.5) / 6 = 20 / 6 = 3.33. Adjust minus 0.25: a broken certificate on the domain the App Store links to is an active, visible operational failure. Base 3.08.

Formula

3.08 Base + 1.00 EV ($150,000, the $0-$1M tier) + 1.00 Geo (USA) + 0.00 Chain = COMPANY 5.08.

For Daxos

PASS. Harmonic shows currently_raising null, one angel round of $150,000 closed 2021-01-11, nothing since, and no valuation figure. There is no round, and no case for creating one: a fifteen-year-old one-time-purchase app in the most commoditized wedge in this category, five people, and a website that does not load cleanly. A check buys a maintained legacy asset, not a seed company. Daxos adds nothing: the CEO's network is enterprise security, and there is no growth motion to accelerate. Fit 1.0 out of 10, the lowest in this batch.

Threat to Moe: 1. No valuation, no listing, no marketplace integration.

What could not be checked

Revenue, install base, cloud-SKU conversion and runway are undisclosed. The CEO's full surname could not be resolved, and the FALSE_POSITION flag makes even the leadership record weaker than the rest of this batch. No co-founder is identified, and the 2021 angel investor is unnamed. Google Play presence could not be confirmed because the site does not load without bypassing certificate validation. The 2020-versus-2011 founding conflict is unresolved.

back to the ranked table

A competent bootstrapped European cross-lister that treats AI as a $4.99 add-on, competes on comparison-page SEO against six near-identical rivals, and does not say publicly who runs it.

What it is

A cross-listing tool for resellers. "Crosslist grows your sales by getting your listings in front of more buyers on more marketplaces, with no extra effort" (Harmonic external_description, pulled 2026-09-16). The site claims one-click integrations with 11-plus marketplaces, inventory transfer between them, CSV import and export, autodelisting, sales analytics, size charts and fee calculators (crosslist.com, 2026-09-16). Harmonic classifies it Business (B2B). The corporate entity is Crosslist BV, per the site footer.

Founders and team

This is the weakest disclosure in the batch and the finding is worth stating plainly.

Harmonic lists two FOUNDER records for this company, person 49011 and person 49118. Both return the same name, Joanne Moore, both are empty of location, education and any experience other than "Co-Founder" at Crosslist with no start date. Two identical empty duplicates is a data artifact, not a team. Harmonic cannot identify who founded this company.

Press coverage names the founders as Gilles Couvreur and Mathias Benoit, with the company headquartered in Kortrijk (GlobeNewswire release, 2026-04-28, carried by Yahoo Finance, Manila Times and Dealroom). That is a paid newswire distribution, so treat it as company-sourced rather than independently reported, but it is the only naming available.

Gilles Couvreur is the one person with a real record. Harmonic person 197296512 lists him as Managing Partner at Crosslist from 2023-02, not as a founder, which conflicts with the press framing. KU Leuven Master of Science plus a postgraduate degree in tax law from Fiscale Hogeschool Brussel. Career before Crosslist: Pricing Analyst at TVH from 2018-10, rising to Senior Pricing Analyst, Tools and Analytics, through 2024-05. He held that full-time role for more than a year after joining Crosslist, so he was part-time on this company until at least mid-2024. He is an analyst, not an engineer.

Mathias Benoit has no Harmonic person record that surfaced and no presence on the site. His background, technical or otherwise, is unverified.

The company site has no about page. The URL crosslist.com/about returns a 404, checked 2026-09-16. No founder or team is named anywhere on the pages fetched. For a company selling a subscription that requires marketplace credentials, publishing no team is a choice.

Checks run: Harmonic company record and three person records, company home, pricing and about pages, US App Store search, one general web search on the founders and the entity. No litigation, dissolved entity or disputed claim surfaced. No adverse record found in the checks run. Two soft flags rather than adverse findings: the missing team page, and a 2026-04 newswire headlined "Crosslist Introduces Cross-Listing Platform" for a company Harmonic dates to 2020-08, which is generous framing of a product that had already existed for five and a half years.

Product and traction

US App Store id 6756124351, checked 2026-09-16: 292 ratings, 4.70 average, released 2026-04-09, last updated 2026-09-05. A Google Play app also exists, com.crosslist.app, linked from the home page, with no published numbers. The site headline claims "4.5 | 1000 reviews," which is an aggregate across sources and does not reconcile with the 292 at 4.70 on Apple.

The honest reading of the traction: six years old, seven staff, zero outside funding, and a mobile app that only launched five months ago. Crosslist has been a browser product for most of its life and is late to mobile in a market where mobile is the whole consumer wedge. Against SellRaze's 39,656 ratings and Hero's 2,060, 292 is a rounding error, though the app is new enough that the comparison is unfair on timing.

The site's own claimed customer outcomes are unsourced marketing: "~156 percent more sales," "80 percent faster," "15-plus hours a week saved." No methodology given.

Business model and unit economics

Four tiers, metered by new listings added per month rather than by crossposts. Bronze $29.99 for 200 new listings, Silver $34.99 for 500, Gold $39.99 for 1,000, Diamond $44.99 for unlimited. Once a listing is in inventory, crossposting is unlimited. AI is not included in any tier. It is a $4.99 a month add-on granting 200 to 2,000 AI credits depending on plan, usable for listing creation and photo editing (crosslist.com/pricing, 2026-09-16). Quarterly billing takes 10 percent off, annual gives two months free. A 3-day money-back guarantee replaces a long free trial, and the page pitches the subscription as 100 percent tax deductible.

The economics are sound and unambitious. Charging AI separately at $4.99 keeps inference cost fully covered and margin clean, and metering on listings added rather than scans avoids the exposure Hero carries. Six years of self-funding to seven staff confirms it works. It also tells you exactly how this company sees AI: a paid feature, not the product.

Position in this market

Crosslist is a follower in the most crowded lane of a category with a hard ceiling. Its own site footer carries six competitor comparison pages: versus Vendoo, Nifty, List Perfectly, PrimeLister, Flyp and Zipsale. A company whose acquisition strategy is comparison-page SEO against six near-identical products is telling you there is no differentiation to defend. Two of those six are in the graveyard or the free tier already.

Where it does differ is geography. This is the only European company in the batch, which gives it some distance from the US-centric competitive crush and access to European marketplaces. It does not change the ceiling.

Platform dependence is the same as everywhere: no public Facebook Marketplace C2C listing API has ever existed, Mercari broke cross-listers in 2024-08, Poshmark suspended cross-lister users twice in 2025, and eBay banned LLM-driven agentic bots in its 2026-02 user agreement. A seven-person team with no burn survives those shocks better than a funded competitor, which is the only structural advantage bootstrapping confers here.

Base decomposition

Market and competition 2.75: the hardest lane in a capped category, entered as a follower, competing on SEO against six named rivals. Product and wedge 4.0: 11-plus marketplaces, both mobile platforms shipped, competent feature set, nothing proprietary, AI bolted on as a $4.99 upsell. Traction 4.0: seven staff self-funded for six years is real, but 292 iOS ratings and an unreconciled "1000 reviews" claim are thin evidence. Unit economics 5.5: clean metering, AI cost fully passed through, 3-day guarantee instead of a long trial, six years of self-funding as proof. Team 3.0: the only identified leader is a pricing analyst with a tax law postgrad who held a full-time job elsewhere until 2024-05; the second founder is unverified; no engineering leadership identified. Founder trust and disclosure 3.5: 404 on the about page, no team named on the site, unusable founder records in Harmonic, unsourced performance claims, and newswire framing that presents a 2020 product as new in 2026.

Mean = (2.75 + 4.0 + 4.0 + 5.5 + 3.0 + 3.5) / 6 = 22.75 / 6 = 3.79. No adjustment applied. The six components already carry the opacity and the follower position, and adding a further penalty would double-count.

The formula line

COMPANY = 3.79 Base + 1.00 EV + 0.25 Geo + 0.00 Chain = 5.04

For Daxos, price and verdict

There is no round and no reason to expect one. Zero rounds in six years, zero recorded investors, Harmonic stage VENTURE_UNKNOWN. This is a small European software business that pays its people and has never sought outside capital.

If it did raise, the case against writing is straightforward. A $100k to $500k cheque into a seven-person follower product in the most commoditised lane of a category whose best exit is a mid-eight-figure tool acquisition cannot return any meaningful part of a fund. The European location adds a further discount on exit routes, since the plausible acquirers here are US marketplaces and US tool consolidators. Daxos has no European reseller-software network to contribute and nothing else this team needs.

The disclosure picture would stop the process anyway. Before any diligence could proceed, Daxos would need to know who actually owns and runs Crosslist BV, whether Couvreur was full-time before 2024-05, what Benoit does, and how a product live since 2020 came to be announced as new in 2026. None of that is available from public sources.

VERDICT: PASS. Not close. File it as evidence that the cross-lister lane is now commoditised enough to support a seventh entrant selling on comparison-page SEO.

What could not be checked

Revenue, subscriber count, churn and profitability: no source. The identity, role and technical background of Mathias Benoit. Whether Gilles Couvreur is a founder or a later-joining partner, given Harmonic and the press conflict. Ownership of Crosslist BV. Google Play ratings and installs. The basis for the "1000 reviews," "156 percent more sales" and "15-plus hours a week" claims. Which of the 11-plus marketplace integrations use official APIs versus browser automation.

back to the ranked table

A one-person Danish app studio that built the best-distributed direct competitor in this wedge, about 22,500 iOS ratings in 15 months, and states publicly that it does not chase funding rounds.

What it is

ThriftAI is a free iPhone and Android scanner for thrifters and resellers. Photograph an item in a shop or at an estate sale and it returns brand and condition detection, an estimated market value against original retail, recent comparable sold prices, a sell-through rate, a match and confidence score, a profit calculator, an authenticity flag, and a Buy or Skip verdict. It then generates a listing and posts to eBay in one tap (thriftai.io and App Store description id 6746565278, read 2026-09-16).

The site names Facebook, Depop, Poshmark, Mercari and Grailed as the price sources. It reads comps from them, it does not write listings to them, so it does not carry the cross-lister enforcement exposure described in the category brief. eBay is the only write target.

Founders and team

Simon Maribo, founder, via simon@thriftai.io and simon@smallstack.dk and a public LinkedIn at linkedin.com/in/simonmaribo. The operating entity is smallstack ApS.

Checks run: Harmonic enrich on thriftai.io and on smallstack.dk, the Danish CVR register via cvrapi.dk on the CVR number printed in the smallstack footer, the studio site root plus /career, /consulting and /about, the App Store lookup and five foreign storefronts, and a web search on the founder and entity. I did not need the Apify LinkedIn scraper because the registry gave verifiable entity facts directly.

The registry confirms smallstack ApS, CVR 45776026, registered 2025-07-24, address Hovmosevej 10, 3400 Hillerod, industry code 582900 "other software publishing", one employee, not bankrupt (cvrapi.dk, pulled 2026-09-16). The app predates the company by seven weeks, so this started as a personal project and was incorporated once it worked. No adverse record found in the checks run: no bankruptcy flag, no dissolution, no litigation and no disputed claim surfaced.

smallstack describes itself openly as a "Modern App Studio" with "1M+ downloads, 4.8 stars across 40k+ ratings, +135 funnels and paywalls built, 9 apps launched and scaled", sells the same skills as consulting (React Native build, Meta and TikTok and Google paid acquisition, paywall and onboarding conversion experiments), and is hiring a full stack developer and a UGC and creator manager. Its careers page says plainly: "We don't take client briefs and we don't chase funding rounds."

Product and traction

About 22,470 iOS ratings at roughly 4.77 across six storefronts, from a 2025-06-05 release, checked 2026-09-16. Google Play package com.fulcra.thriftai shows 250,000-plus downloads (search result, 2026-09-16). Version 1.0.78 dated 2026-09-02, which is a very high release cadence for 15 months.

Measured on rating count per month live, this is the fastest consumer ramp in the whole batch: roughly 1,500 iOS ratings a month, against Curio's 650 and Value Scout's 90.

The competitive picture is equally stark. A single App Store query for "thrift ai" returned Thrifty (378 ratings), Thrift AI Profit Identifier (154), Thrift AI Profit Finder Cluzy (190), The Thrift AI Profit App (100), ThriftLens (161), Thrift AI Resale Profit ID (0), ThriftHaul (40) and SnapFlip (22), all released after ThriftAI. Eight named clones of one product, inside 15 months.

Business model and unit economics

Free three-day trial, then a mandatory subscription, stated in the first two lines of the App Store description. That is the highest-converting and highest-churning paywall shape in consumer mobile, run by an operator whose own marketing lists "135 funnels and paywalls built" as a core competency. The site carries a dedicated /terms/refund-policy page, which usually means refund volume is high enough to need one.

Revenue is real and probably good for one person. It is subscription cash on a use case the user needs for a few weekends, not a compounding account base, and there is no take rate on the eBay listings it generates.

Position in this market

ThriftAI is the most credible operator in the consumer scan-value-list wedge in this batch: a named founder, a registered company, verifiable registry data, 250,000 downloads and weekly shipping. It also demonstrates the category ceiling more clearly than anything else here. One person with no capital built the best-distributed product in the lane, eight people cloned it within a year, and the founder's own public position is that funding rounds are not the point. That is what a commodity wedge looks like from the inside.

Base decomposition

• Market and competition 2.0. Commoditized consumer wedge, free inside eBay, Facebook Marketplace and Google Lens, with eight named App Store clones of this exact product shipped in 15 months.
• Product and wedge 3.5. Genuinely better built than its peers (sold prices, sell-through rate, profit calculator, authenticity flag, one-tap eBay listing) and read-only on the risky marketplaces, but nothing defensible.
• Traction 6.0. About 22,470 iOS ratings at 4.77 and 250,000 Play downloads in 15 months, the fastest ramp in the batch.
• Unit economics 4.0. Hard three-day-trial paywall converts well and churns hard, with a dedicated refund policy page and no transaction revenue.
• Team 3.0. One verifiable, technical, unusually effective founder, spread across a nine-app studio with one registered employee.
• Founder trust and disclosure 5.0. Best in the batch: named founder, public LinkedIn, verifiable CVR, real studio site, published refund policy, and an honest public statement about not raising. Minor knock for a "32K+ ratings" homepage claim above the roughly 22,500 I could count on iOS, and for homepage counters that render as zeroes.

Mean = (2.0 + 3.5 + 6.0 + 4.0 + 3.0 + 5.0) / 6 = 23.5 / 6 = 3.92. Adjusted -0.17 because the cloning intensity is the dominant fact about this asset and will compress it. Base = 3.75.

Formula

3.75 Base + 1.00 EV ($0 raised) + 0.25 Geo (Denmark, EU) + 0.00 Chain = COMPANY 5.00.

For Daxos

Pass, and the company has already said so itself. smallstack's careers page states "we don't chase funding rounds", there is no round, no cap table and no raise, and the entity is a one-employee Danish ApS running nine apps rather than a company building one. Even setting willingness aside, a $100k to $500k check into a subscription scanner in a lane with eight live clones cannot return a seed fund, and Daxos adds nothing to Danish consumer mobile growth. DAXOS 2.0, VERDICT PASS.

The useful output is competitive intelligence, not a deal. Simon Maribo is the clearest existence proof in this research that the consumer wedge can be reached by one person with no capital, which is the strongest argument against funding anyone else to try it.

What could not be checked

Revenue, MRR, subscription price points, trial conversion rate, refund rate, churn, and paid acquisition spend. The CVR record has a protected address and returns a null owners field, so beneficial ownership was not confirmed from the registry, only the entity itself. The 32,000-rating and 1M-download claims are the company's own and were not independently verified beyond the storefront counts above. I did not open the LinkedIn profile directly.

back to the ranked table

The best consumer traction per dollar in this batch, 627 ratings at 4.83 on zero raised, but the app has not shipped in nine months and the founder started a second company the same month it went quiet.

What it is

A direct Moe competitor and the purest example of the commoditised wedge in this set. Point the camera at an object, get an identification, a valuation based on market factors and condition, links to similar items for sale, and a listing (whatsit.ai, fetched 2026-09-16). App Store positioning is thrifting: "Use WhatsitAI at yard sales, thrift stores, or even in your attic. Find out if that vase is a rare masterpiece or if that figurine is a collectible."

There is a second, grander story in the Harmonic record: "building the intelligence layer for physical assets. Trillions of dollars of physical items exist without a system of record. The same object gets identified and valued repeatedly across resale, insurance, restoration, and commerce, often with different answers each time." Nothing on the live site or in the app supports that B2B framing. It reads as investor-deck language attached to a consumer thrifting app.

Founders and team

Don Milley, Founder and President, start 2024-04. Cornell BS, dual Business and Animal Physiology. Not technical: no engineering degree, no engineering title anywhere in the record, no GitHub surfaced.

Career from Harmonic: Lead Consultant in digital product management and M&A at Acquisition Associates 2003 to 2018; VP of Product and Customer Engagement at Alphabuyer 2014 to 2016; Director of Product and Program Management at Piano 2018 to 2019; Interim Director of Product Management at Riteload 2019; Senior Digital Transformation Consultant at Kforce 2019 to 2021; Head of Product at PEAC Solutions 2021 to 2022; Senior Director of Product Management at Corcentric 2022 to 2024. A twenty-year product management career. Three prior ventures, none with a visible outcome: iParents.com 2008 to 2010, PerDiemDeals.com LLC 2010 to 2011, Groupinion 2012 to 2015. Serial founder, no exits on record.

The finding that matters: Harmonic shows Milley as Founder and CEO of ContentsIQ from 2025-12-01, current, and Entrepreneur in Residence at Drexel University's Office of International Programs from 2025-09-01, current, both alongside WhatsitAI. ContentsIQ, by Claims Crafters, is AI-powered residential contents valuation for insurance claims (claimscrafters.com and Restoration and Remediation Magazine, read 2026-09-16). The WhatsitAI app's last update is 2025-12-22, the same month ContentsIQ started. That correlation is the clearest read on where his attention went.

Kohei Hashimoto, Senior Technical Lead, start 2025-10, located in Japan. MS Computer Science in machine learning from UMass Amherst, BA in artificial intelligence from Nagano University. Prior: iOS developer at lgcy 2022 to 2024, full stack engineer at Tweebaa 2025-02 to 2025-11, which overlaps his WhatsitAI start by a month. Genuinely technical, but remote and hired eighteen months after founding.

Matthew Martinez, Data Analytics and AI Modeling Product Owner, start 2024-08, still listed as current. Lehigh undergraduate in Computer Science and Business, co-founder of Finteera, and since 2026-06 a business analyst intern at Veolia. A student, not a full-time operator.

That is three of the five recorded people: a founder running a second company, a remote technical lead, and a college student.

Checks run: Harmonic company record and three person records, company site, App Store listing and description, web search on the founder, the company and ContentsIQ. No adverse record found in the checks run.

Product and traction

627 ratings at 4.83, released 2024-10-01, updated 2025-12-22 at v2.1.1 (iTunes, 2026-09-16). All 627 are against the current version.

Per dollar that is the best consumer traction in this batch: 627 ratings at 4.83 on zero recorded funding beats SellOut's 152 on zero, Hamelyn's 214 on $4.7M and Speeral's 0 on $1.8M. Users who use it like it.

Against the live benchmark it is still small: SellRaze has 39,656. And the store is full of near-identical clones: RelicSnap 709 ratings, AntiqueScan 110, Furniture Identifier 52, Antique Doll Appraisal 32, most solo-developer apps. That is what a commodity wedge looks like from the inside.

Harmonic web_traffic 4,283. LinkedIn 138 followers, Facebook 137. Nine months with no shipped update.

Business model and unit economics

From the App Store description: a free plan with five ad-supported scans per week, then a paid upgrade for unlimited searches and no ads. Prices are not published.

The economics are poor and structural. Every scan is a vision-model inference that costs real money, so free users are gross-margin negative, and ad revenue against five scans a week from a base this size is negligible. Conversion to a consumer subscription is hard when Google Lens, Apple Visual Intelligence and free-tier ChatGPT answer "what is this and what is it worth" free on the same phone.

Position in this market

The consumer scan-and-value leg with nothing else attached, which the brief identifies as the one thing the platforms and the operating systems have already absorbed. eBay, Facebook Marketplace, Depop, Mercari and Poshmark all ship photo-to-listing with suggested prices free, and the OS layer answers the valuation question free. There is no wedge left at this altitude and no path from a thrifting companion to a demand network.

The founder appears to have reached the same conclusion. ContentsIQ points the identical engine at insurance claims adjusters, who have a budget, a compliance requirement and no free alternative. That is the correct move per the brief: go vertical, go where comps are needed and money already changes hands. It is also a different company.

Base decomposition

• Market and competition 2.5: the most commoditised function in this category, free at the OS and search layer, with a swarm of solo-developer clones on the same shelf.
• Product and wedge 3.5: a complete and well-liked product for a real behaviour (sourcing at thrift stores), but the wedge is a feature.
• Traction 4.0: 627 ratings at 4.83 on zero dollars is the best efficiency in this batch, discounted hard for nine months without a release.
• Unit economics 3.0: inference cost on every free scan, negligible ad revenue, a consumer subscription competing against free.
• Team 3.0: a non-technical serial founder with no exits now running a second company, a remote technical lead hired at month eighteen, an undergraduate.
• Founder trust and disclosure 3.5: the "intelligence layer for physical assets" framing is far larger than anything shipped; the site's income testimonial is properly disclaimed; the divided attention is nowhere disclosed.

Mean = (2.5 + 3.5 + 4.0 + 3.0 + 3.0 + 3.5) / 6 = 19.5 / 6 = 3.25. Adjust -0.25 because the product has not shipped since the month the founder started ContentsIQ, which makes this a business being run down rather than built. Base = 3.00.

Formula

3.00 + 1.00 (EV, Harmonic funding_total $0, in the $0 to $1M band) + 1.00 (Geo, USA) + 0.00 (Chain) = 5.00.

As with SellOut, the EV term flatters. WhatsitAI has a better claim to it than most, since 627 ratings at 4.83 on nothing is genuine. But the Base of 3.00 is the number that describes the business.

For Daxos, price and verdict

Harmonic shows currently_raising null, zero rounds, no valuation and no valuation_info, so there is no price on the table. The live benchmark for this wedge is SellRaze at $500k on a Harmonic-estimated $2.0M post, and SellRaze has sixty times the ratings. WhatsitAI would have to price well below that, and at any price Daxos would be buying a stale consumer app whose founder's main job since December 2025 is elsewhere, with a check that cannot return a fund in a category whose tool lane has a proven hard ceiling. DAXOS = 2.5. VERDICT PASS.

One note to carry forward rather than act on. Milley has already made the move the brief says is the only one that works: same engine, vertical buyer, liquid comps, budget on the other side. That vehicle is ContentsIQ by Claims Crafters, a separate entity, and it is the thing that would justify a conversation. Backing WhatsitAI gets Daxos no exposure to it.

Threat to Moe: 1. Same wedge and country, unshipped for nine months with an absentee founder.

What could not be checked

Subscription pricing, paying subscribers, revenue, scan volume. Whether the "Investor and Advisor" listed from 2024-09 put in money: Harmonic records zero rounds and an empty investor list, so no web figure was substituted. Google Play presence, not linked from the site. The ownership relationship between Milley, WhatsitAI, Inc. and Claims Crafters, and whether WhatsitAI technology was licensed or assigned to ContentsIQ. Whether the B2B positioning in Harmonic reflects a real pipeline or deck language. LinkedIn was not scraped.

back to the ranked table

Four Berkeley finance graduates building securitized index funds of graded trading cards, with the sharpest original research in this scan, no engineer, no product, and a business model that has already failed three times.

What it is

Not a resale tool. tash is building a brokerage for trading cards: a registered investment vehicle holding a diversified portfolio of fifty graded, investment-grade cards, with shares in that vehicle bought and sold on tash's own platform. The framing on the homepage is explicit: "A MARKETPLACE SELLS YOU MERCHANDISE. A BROKERAGE ISSUES YOU OWNERSHIP", with vaulted custody and SEC oversight. Y Combinator's directory entry calls it "The BlackRock of collectibles" (https://www.ycombinator.com/companies/tash, read 2026-09-16).

Nothing has launched. The site is a waitlist, and the YC entry states "We are currently pursuing SEC qualification to launch regulated investment offerings backed by trading cards." Source: https://tash.cards, read 2026-09-16.

Founders and team

Four co-founders, all University of California, Berkeley, class of 2026, incorporated as Tash Cards, Inc. in January 2026. Three of the four, Shammas, Ohannesian and Borshell, also attended Loyola High School of Los Angeles between 2018 and 2022. This is a high school and university friend group.

John Ohannesian, Founder/CEO (Harmonic person 195074082). Indiana University Kelley School 2022 to 2024, then Berkeley 2024 to 2026. Houlihan Lokey investment banking summer analyst 2025, and an incoming Bank of America investment banking seat dated 2025-12 to 2026-05 that he left to do this.

Nathan Borshell, Founder/CTO (Harmonic person 197727190). Economics and Data Science at Berkeley. UBS Global Markets summer analyst 2025, and an incoming Barclays investment banking summer analyst position dated 2025-09 to 2026-05. He has never held an engineering role on any record checked, and no GitHub or shipped software surfaced. A CTO title on a markets analyst is the single most important fact on this card.

Victor Shammas, Founder/CPO (Harmonic persons 163798972 and 196372228, duplicate records). Berkeley 2022 to 2026 with a 3.97 listed, Haas. Summer investment analyst at FPA Funds 2024, data analyst at Seso, intern at Galliott Capital Advisors.

Nicolas Sharma (Harmonic person 180460374). Berkeley Data Science, Economics and Haas Business degrees. J.P. Morgan investment banking summer analyst 2025, data scientist at Capgemini summer 2024. The nearest thing to a technical person on the team, on the strength of one summer.

So: four capable finance students with real credentials in markets and none in building software, proposing to build custody, compliant share issuance, an order book and a secondary market. Harmonic also records five founder relationships at a four-person company with start dates of 2026-02, 2026-02, 2026-04, 2026-04 and 2026-05, which says the team was still assembling during the batch.

Checks run: Harmonic company enrich on tash.cards, Harmonic person records for all five founder urns, the company site including the research page, the Y Combinator company page, and web search on the fractional collectibles category history. No adverse record found in the checks run: no litigation, dissolved entity, regulatory action or disputed claim against these individuals or this entity surfaced. The Apify LinkedIn actor was not run because the Harmonic person records already carried degrees, employers and dates.

Product and traction

No product, no revenue, no users. Waitlist only, pending SEC qualification, seven months after incorporation.

What exists is distribution and thinking. Harmonic puts web traffic at 20,250, which is high for a pre-launch page and reflects the YC brand plus one genuinely good piece of writing. The research page carries "Sector Thesis No. 1: The Census Asset" (Tash Cards Inc., July 2026), arguing that graded cards are the only passion asset whose supply is an auditable public ledger rather than an assertion, because every card certified since 1991 sits in a queryable population report by issue and grade. It then argues that this, not nostalgia, is why index construction is possible in cards where it failed in art, and it says bluntly that the industry's own promotional statistics are now the main obstacle to institutional adoption. That is real original analysis and the best piece of work produced by anyone in this scan.

Against it, the homepage hero. The chart headed "Quietly outperforming the S&P 500. For two decades", showing a CL50 index at $23,632 against the S&P at $10,163, carries a note left visible on the live site: "CL50 curve anchored to documented values ... Intermediate points interpolated; final chart to be generated from the Card Ladder daily series." The most persuasive claim on a site marketing a securities product is an unfinished placeholder, shipped with the internal build note attached.

Business model and unit economics

Undisclosed. The structure implies a management fee on assets plus a spread on sourcing, against fixed costs that are unusually heavy: Reg A+ qualification, ongoing reporting, a transfer agent, a broker-dealer relationship, audit, insured vault custody and grading. None of that scales down. $500,000 does not fund a securities platform to qualification and launch, so this company must raise again soon and on terms set before it has a single dollar of AUM.

Position in this market

tash is adjacent to the resale scan, not in it. It never touches a listing, a marketplace API or a seller. It is exposed to none of the platform risk in the brief, and it threatens Moe not at all.

Its own category history is the problem, and it is worse than the resale graveyard. Every prior attempt at fractional collectible ownership has failed. Otis was acquired by Public.com in March 2022 and its single-asset series were wound down and liquidated through 2023 and 2024, some at steep losses. Collectable stopped running new offerings and ended in Delaware Chancery Court litigation. Rally's parent RSE Markets paid $350,000 to settle an SEC enforcement matter in 2023, disclosed a going-concern warning on its 2025 audit, and has announced it will close the platform and roll remaining assets into an NYSE-listed entity called PIKA. Sports Illustrated's headline on the sector reads "Fractional Investing in Sports Cards Has Been A Failure" (https://www.si.com/collectibles/news/fractional-investing-in-sports-cards-has-been-a-failure). Sources also https://alts.co/lets-mourn-rally/, read 2026-09-16.

tash has one specific answer to that history, and it is the right one to have: an index of fifty cards rather than a single-asset series. Single-asset series are what made Otis and Collectable illiquid and unsellable. Diversification plus a real secondary market is the correct diagnosis. Whether four people with $500,000 and no engineer can execute it where funded teams failed is the question.

Base decomposition

• Market and competition: 3. A genuine asset class at $500M to $700M of monthly volume by their own chart, and the census argument for underwritability is sound. But the specific business model is zero for three, and the incumbent for card-backed financial products, Alt, has raised $346M.
• Product and wedge: 3. Index-of-fifty inside a registered vehicle is a real structural fix for the illiquidity that killed the predecessors. It is also entirely a slide: nothing is built, nothing is qualified, nothing is live.
• Traction: 2. No product, no revenue, no AUM. 20,250 web visits and 2,799 LinkedIn followers pre-launch are good, but they measure the YC brand and one essay, not demand for the security.
• Unit economics and business model: 2. Undisclosed, and the comparable cost structure has bankrupted better-capitalised operators. Fixed regulatory and custody costs do not scale down to a $500,000 company.
• Team: 2. Four finance students with strong banking internships and no engineer. The person titled CTO has never worked in engineering. They must build custody, issuance, an order book and a compliant secondary market.
• Founder trust and disclosure: 4. Honest about pursuing rather than holding SEC qualification, honest about past performance, and openly critical of the industry's own promotional numbers, which is rare and good. Pulled down by shipping an interpolated performance chart with the placeholder note still on it as the hero claim of a securities marketing page.

Mean = (3 + 3 + 2 + 2 + 2 + 4) / 6 = 16 / 6 = 2.67. Adjusted up 0.25 to 2.92, for the two things that genuinely distinguish this from a generic YC pre-launch: the index structure is a specific and correct answer to why the predecessors died, and the census thesis is the sharpest original analysis in the whole scan.

The formula

Base 2.92 + EV 1.00 ($500,000 raised, $0 to $1M tier) + Geo 1.00 (San Francisco) + Chain 0.00 = COMPANY 4.92.

For Daxos

Pass, on price and on model.

Price first. This is an active Y Combinator Summer 2026 company that has taken the standard $500,000 and will raise a post-batch seed on YC demo day terms. Those rounds price at roughly $15M to $25M post for a pre-launch consumer fintech with a good story, and this company has a good story. At that price a $100k to $500k check buys between half a percent and three percent, with no information rights worth having and no ability to shape the outcome. Daxos brings nothing a YC seed round needs.

Model second, and this is the binding objection. Fractional collectible ownership has been tried at scale three times with far more capital and it failed three times, once ending in an SEC settlement and a going-concern warning. tash's index structure is the right correction, but the team executing it is four people who graduated this year, the one titled CTO has never shipped software, and the hardest parts of the build are custody, compliance and market-making rather than a marketing site. The $500,000 does not reach qualification.

What would change this: a live Reg A+ qualification, a real engineering hire with brokerage or custody experience, and early secondary-market volume showing shares actually trade. If all three happened and the price were still under roughly $12M post, it would be a MEET. Today it is neither reachable nor ready. DAXOS 2.5. VERDICT PASS.

What could not be checked

Fee structure, index methodology beyond the fifty-card description, the custody and grading partners, which vehicle structure they are pursuing (Reg A+ tier, Reg D, or series LLC), and how far along SEC qualification actually is. Waitlist size is not published. Whether the $500,000 is solely the standard YC investment or includes other money could not be separated from the single Harmonic round. No revenue, AUM or cost figures exist to check. Whether any engineering hire has been made since the Harmonic headcount of 4 is unknown.

back to the ranked table

The best founding team and the best domain in this batch, $5M from Thiel and crypto funds in June 2023 on a Web3 thesis it has since dropped, and three years later a mid-pack app and an empty community feed.

Scope check

Harmonic funding_total on Collectbase Inc. is $5,000,000, below the $10.9M threshold. In scope, rated below.

What it is

Collectibles.com is a scan, identify, value and collection-management product covering sports cards, Pokemon, coins, comics, stamps, dolls and vintage toys, with a community feed, a leaderboard and a "Repackz" section (collectibles.com, fetched 2026-09-16). The same company runs a second property, Cardbase at getcardbase.com, which aggregates trading card prices from "over 30 marketplaces / auction houses" and indexes roughly 144 million cards by its own category counters (54.8M baseball, 34.1M football, 24.1M basketball, 10.4M Pokemon and so on, read 2026-09-16).

Harmonic's stored description still reads "Building the world's first Web3 community + marketplace for collectors and the global collectibles industry". Nothing on either live site mentions crypto, tokens or Web3 today. The money was raised on a thesis the company no longer runs.

Founders and team

Alexander (Oleksandr) Ivanov, CEO of Collectbase since January 2021 and co-founder of Collectibles.com since September 2021, based in Dubai (linkedin.com/in/ivanovoleksandr). MSc and BSc in Business Information Systems from Göttingen and Paderborn. Previously co-founder and CPO of Buchdeals (2018 to 2020), CPO at Tatar Holding (2017 to 2018), and co-founder and CPO of Rivalfox (2013 to 2016).

Anas Salem, co-founder since October 2020 and COO of Collectibles.com since January 2023. Dipl.-Wirt.-Ing in industrial engineering from RWTH Aachen. Co-founder with Ivanov at both Buchdeals and Rivalfox.

Dietrich Von Behren, angel investor and advisor from August 2021, then co-founder and Chief Business and Strategy Officer from September 2022 (linkedin.com/in/vonbehren). Previously SVP strategy and business development then Chief Business Officer at Reelgood from 2017 to 2022, a Series A company that has raised $15.35M.

Checks run: Harmonic enrich on collectibles.com, collectbase.com and collectbase.io, Harmonic person records for all three founders (ids 347912, 157516106, 164576195), Harmonic company records for four prior employers (ids 10586381, 1545036, 56108091, 767981), both live sites, the App Store lookup across five storefronts, and the Danish-style registry check did not apply here. I did not run the Apify LinkedIn scraper because Harmonic returned full structured experience for every founder.

Two findings to record honestly rather than as scandal. Rivalfox, the venture Ivanov and Salem co-founded in 2013, raised $150,000 and Harmonic records it as OUT_OF_BUSINESS. Ivanov and Salem have now been co-founders together across three ventures and 13 years, which cuts both ways: durable partnership, no exit yet. Von Behren's LinkedIn headline claims "4x Founder, 2x Exits", which I could not verify against the record; his listed Reelgood roles are executive rather than founding, and the headline presents them as prior employers rather than as founded companies, so nothing there is misstated. No litigation, dissolved entity, bankruptcy or disputed claim surfaced for any of the three.

The gap is engineering. No CTO or lead engineer appears in Harmonic, on either site, or in search. The only non-founder employee Harmonic records is a product designer in Warsaw. Ivanov is a three-time CPO with an information systems degree, which is product-technical, but there is no shipped-code or repository evidence, so I record TECHNICAL_FOUNDER as unknown rather than yes.

Product and traction

About 10,590 iOS ratings at roughly 4.6 across five storefronts, from a 2020-10-07 release, last updated 2026-09-01 at v4.4.1. The app is six years old and actively maintained.

The number is weak for its age in a crowded field. From the same App Store queries on 2026-09-16: Collectr 64,990 ratings, PSA Card Scanner 56,837, CollX 48,085, Sports Card Investor 37,718, PriceCharting 26,463, HoloDex 22,673. Collectibles.com is the oldest app in that list, has a premium domain and $5M, and sits sixth.

The community feed on the homepage reads "Hey! Looks like no posts exist yet. Be the first to create a post!" as of 2026-09-16. LinkedIn follower counts are 545 for collectibles-com and 3,174 for collectbase; the Twitter account is 798 followers.

Business model and unit economics

Affiliate commissions are the only disclosed revenue: getcardbase.com carries an explicit eBay Partner Network disclosure. Collection tracking is advertised as free. No subscription price, no take rate and no marketplace fee is published on either property.

Six years of operation and $5M consumed, with the last round in June 2023, against an affiliate-only disclosed model and no revenue figure anywhere.

Position in this market

This sits in the vertical-with-liquid-comps corner the category brief identifies as the only place the thesis compounded, and it is the most crowded niche in the whole research set. The company has the best brand asset in the batch, the domain collectibles.com, and a founding team with more relevant depth than anyone else here. It also raised on a Web3 marketplace thesis from crypto funds in June 2023, abandoned that thesis, and has since been out-distributed on the app by a four-person bootstrapped competitor and by a one-person Danish studio in an adjacent wedge.

Base decomposition

• Market and competition 3.5. Right vertical, wrong density. Six better-distributed apps in the same niche, several of them free and better funded.
• Product and wedge 3.5. A premium domain, a genuine 144M-card price index at Cardbase, and a collection manager, against a dropped founding thesis and a community feature that is visibly empty.
• Traction 4.0. About 10,590 ratings at 4.6 over six years and still shipping, but sixth in its own niche and beaten by apps three and five years younger.
• Unit economics 3.0. Affiliate-only disclosed revenue on a free product after $5M and six years. Weakest capital-to-monetization ratio in the batch.
• Team 5.0. Strongest founding team here: two repeat co-founders across 13 years, a serial CPO, and real BD depth from Reelgood. Offset by one prior venture recorded out of business and no identifiable engineering lead.
• Founder trust and disclosure 4.0. Named founders, a stated Delaware corporation, published terms and privacy, real investor names. Offset by $5.0M and $0.00 on two Harmonic records for one entity, headcount reported as both 2 and 11, a stale Web3 description, and no public update since the 2023 round.

Mean = (3.5 + 3.5 + 4.0 + 3.0 + 5.0 + 4.0) / 6 = 23.0 / 6 = 3.83. Adjusted -0.25 because three years and three months past a $5M seed with a mid-pack app, no disclosed revenue and an empty community is the dominant fact. Base = 3.58.

Formula

3.58 Base + 0.25 EV ($5,000,000, the $5M to $10M tier) + 1.00 Geo (USA, Delaware corporation, Mountain View) + 0.00 Chain = COMPANY 4.83.

For Daxos

Pass at the price this would come at. The last round closed 2023-06-13, so the company is 39 months past its seed. That is either quiet profitability or a stall, and an empty community feed, a sixth-place app, affiliate-only monetization and conflicting headcount records all point to stall. The cap table already holds Peter Thiel, Blockchange, GFR Fund, Blockwall and a set of celebrity angels, so a $100k to $500k Daxos check would be immaterial capital joining a crowded table, and Daxos brings nothing to collectibles data that this team does not already have.

One condition would flip this to MEET rather than PASS: a genuine recapitalization at a reset price with disclosed revenue and a named engineering lead. On the evidence available today there is no round to enter and no reason to want one. DAXOS 2.5, VERDICT PASS.

What could not be checked

Revenue, GMV, MAU, runway, and whether any capital was raised after June 2023, since Harmonic shows one round and no valuation field on either record. Headcount is genuinely unresolved: Harmonic's two records for the same operating entity report 2 and 11. Google Play presence for the Collectibles.com app was not confirmed. I could not verify Von Behren's "2x Exits" claim. Whether the Web3 pivot involved any token issuance, and what happened to it, is not visible from public sources.