Aggregate data

AI startup benchmarks

Revenue-retention distributions from early-stage AI startups whose metrics are read directly from Stripe. Aggregate and anonymised — distributions only, never an individual company.

Estimated 30-day gross revenue retention — distribution

No snapshot published yet

Foundr publishes a distribution only once the cohort is large enough that no individual company can be inferred from it, and only as a dated snapshot. The methodology below is complete and applies to every snapshot when one is published.

How these numbers are produced

  • Every figure comes from a startup's own Stripe account, connected by the founder via OAuth and read by API — never typed in.
  • The cohort is startups listed on Foundr whose listing is approved and which have Stripe-verified retention. Connecting Stripe for the private founder benchmark alone does not place a company in a published cohort.
  • Retention here is an estimate computed as current MRR ÷ (current MRR + MRR lost to cancellations in the last 30 days). It is derived from live subscription state, not from a fixed opening cohort, so a fast-growing company's figure reads higher than a textbook gross-retention calculation would give.
  • Snapshots are frozen and dated. Foundr does not continuously republish a live distribution: repeatedly refreshing a distribution over a small population can reveal how an individual company moved, even when no company is named.
  • Distributions only, in coarse bands, with small cells shown as a range rather than an exact count. Individual company figures are never public — they sit behind an invite-gated, accredited-attested login.
  • These statistics describe a population of companies. They are not investment advice and say nothing about any individual company.

Full pipeline and security details: How verification works →

Citing this page

Foundr. “AI startup benchmarks.” Methodology; no snapshot published yet. https://foundr.ventures/benchmarks