ProofStack
What this site is, and how it works
ProofStack is an independently built and operated library of 50 audited small-software business cases. It exists because two things kept being wrong on the internet: revenue screenshots nobody can verify, and AI answers that invent numbers.
Evidence grades
- A — Live, revisitable data: public transparency dashboards you can check any day (e.g. real-time ARR panels).
- B — Dated, founder-signed disclosures: blog posts and interviews written at the time of the events.
- C — Official pages without figures: the product and pricing structure are verified; revenue stays unknown.
Currently 15 of 50 cases carry a dated public revenue disclosure. The rest say "not disclosed" — unknown is not zero, and it never becomes an estimate here.
Honesty rules
- Every third-party claim carries its source URL and an observation date; snapshots are dated and never silently updated.
- Acquisition stories come from first-party sources; cases not yet researched say so instead of showing template text.
- Counts shown on this site are either computed live or explicitly dated — no third state.
- Corrections policy: material errors are fixed in place with a dated note, and the error is recorded in our internal operations ledger rather than quietly deleted. We have corrected our own data before (e.g. a mislabeled primary acquisition channel) and will again.
What runs this site
A single independent operator plus an automated operations system: cases are extracted with AI assistance but pass machine quality gates and human source checks before publishing; benchmarks, digests and indexes are computed from the verified data, not written by hand. The paid service is a $99 human-verified evidence audit — payment after delivery.
Contact & machine access
Email: proofstack@outlook.com · llms.txt · llms-full.txt · MCP · dataset (CC BY 4.0) · RSS