2,342 startups for sale, with revenue read from the payment processor
For anyone evaluating the purchase of a small software business. Asking price, revenue actually collected, two computed multiples, and a flag on every listing whose displayed figure has stopped updating.
The proof, before anything else
Ten rows straight out of the file. Nothing here is an illustration. Revenue comes from each company's payment provider API, not from a seller's claim.
| Company | Market | Rev 30d | Rev 12m | Price | TTM mult. | Payback |
|---|---|---|---|---|---|---|
| TopDividendTools | Marketplace | 9,878 | 52,718 | 5,000 | 0.09x | 0.5 m |
| Kortex-Notebooklm | AI | 2,436 | 232,363 | 50,000 | 0.22x | 20.5 m |
| Stealth Company | Mobile Apps | 4,970 | 264,363 | 90,000 | 0.34x | 18.1 m |
| Stealth Company | AI | 3,045 | 91,817 | 35,000 | 0.38x | 11.5 m |
| LifeFlexIA | SaaS | 1,606 | 26,596 | 11,500 | 0.43x | 7.2 m |
| DOWNCONVERT SKETCHUP FILES | Productivity | 5,176 | 39,877 | 17,900 | 0.45x | 3.5 m |
| Okfans | Entertainment | 4,307 | 56,331 | 30,000 | 0.53x | 7.0 m |
| LinkPost | AI | 4,859 | 144,945 | 79,000 | 0.55x | 16.3 m |
| SHANNON LAB LLC (ONLY STRIPE MRR C | AI | 10,354 | 36,895 | 21,000 | 0.57x | 2.0 m |
| ColdSire | SaaS | 12,021 | 248,763 | 149,000 | 0.60x | 12.4 m |
The problem
The finding that started this work is published in full, for free, in the stale-connection audit.
TrustMRR shows a leaderboard and profile pages. It shows no multiple, it cannot sort by price relative to revenue, and above all it does not flag an expired payment connection. Across the corpus, 950 profiles display a stopped figure, and those profiles carry 47% of all revenue presented as current. Among the for-sale listings, 62 show frozen revenue while asking $8,130,343 between them.
Comparing thirty listings by hand means opening thirty pages, copying six fields from each, and computing the rest. Comparing all 2,342 is out of reach.
How the snapshot was built
- Read the 12,255 public pages TrustMRR publishes in Markdown for AI agents, at the rate the site itself publishes in its headers, sixty requests a minute. No authenticated access, nothing circumvented.
- Extracted twenty-five fields per company, including two the leaderboard never shows, the state of the payment provider key and the timestamp of the last real reading.
- Computed two separate multiples and a gross payback period, then sorted cheapest-first relative to revenue.
Is the snapshot still accurate
Checked the same day. I re-read twenty-four listings drawn at random from those collecting $1,000 a month or more, straight from the source, ten hours after the collection.
- Asking price and payment connection state matched on every one.
- Median revenue difference was 0.6%, and twenty-three of the twenty-four moved less than ten percent.
- The ranking by multiple barely moved, two ranks changed out of twenty-four, average displacement 0.1 rank.
Displayed revenue covers a rolling thirty-day window, so it moves every day. What this measurement shows is that the daily movement does not change the order, so the snapshot stays usable for what it is for, spotting the listings worth a closer look.
What you get
- A filterable dashboard, one HTML file that opens offline, with no remote script and nothing transmitted. Filter by market, minimum revenue, maximum multiple, maximum price and connection state.
listings.csv, 2,342 rows and twenty-five columns, for a spreadsheet or any tool.- Both multiples per listing, trailing twelve months where history exists (1,802 listings) and last-30-days annualized (1,222 listings), plus gross payback months.
- The stale-connection flag with the date of the last real reading, on every row.
- The context report, success rates by revenue tier, by market and by channel, median age to $1k and to $10k a month, and revenue concentration, computed on the 11,303 companies with a live connection.
Limits, written before you find them
- This is a dated snapshot, not a subscription. The figures are those of 6 November 2026. The check above measures one day of drift, it says nothing about a month.
- Revenue is money collected, not profit. Nothing here measures margin, operating cost or hours worked.
- Asking price, description, channels and team size are stated by the seller. Only revenue is verified by a third party.
- The corpus is self-selected. A company appears only if it connected its payment provider, and there is no way to know how many declined.
- A very low multiple more often signals a problem the profile does not show than a mispriced opportunity. The snapshot orders information. It verifies no listing and recommends no purchase.
- The data comes from a third-party site, read publicly. This is independent analysis, with no affiliation to TrustMRR.
$39 USD
One-time payment, no subscription, no renewal. Delivered by email within a minute. Full refund on request within thirty days, no questions.
Questions
- Why not just read TrustMRR, it is free.
- You can, and you should for any listing you care about. The snapshot does not replace reading a profile. It tells you which ones are worth reading, and saves you opening 2,342 pages to find out.
- Are these the site's own numbers.
- Yes for revenue, asking price and the seller-stated fields, carried over unchanged. Both multiples, the payback period and the stale flag are computed by me from fields the site publishes but does not use.
- Can I verify a row.
- Every row carries the URL of its source page. It is the last column of the CSV.
- What if a number is wrong.
- Email me the row. If the error is in my extraction I fix it and resend the file to everyone who bought it. If it comes from the source, I note it in the file.
- Will there be an update.
- I promise none, because I am not selling a subscription. If I publish one, buyers of this snapshot get it at half price.