benchmarks/company funding/most accurate funding data
funding data · accuracy, measured

Most accurate company funding data

“Most accurate” is the most contested claim in funding data — and it means two different things. Accuracy when present: when a provider returns a stage, how often is it right? Correct stage yield: across every company with a verified stage, how often is the right one returned — penalizing silence. A provider can look accurate by only answering the easy companies; the second metric closes that loophole.

Seven providers, the same 300 identity-verified companies, every stage judged against a primary source. Measured answers below — not a vendor grading its own homework.

most correct end to endFiber77.2%

open data + codeInputs, normalized provider outputs, and evaluation code are public in openbenchmarks-labs/company-funding.github →

Two accuracy answers, both measured

Pick the sense that matches your workflow. If you fill missing stages downstream, weigh accuracy when present; if you consume the stage as-is, weigh correct stage yield.

Accuracy senseWhat it measuresLeader
Correct stage yieldRight latest stage across all companies (silence penalized)Fiber 77.2%
Accuracy when presentRight stage among the ones actually returnedFiber 84.2%
Stage fill rateHow often a provider returns any stage at allFiber 91.8%

Seven providers on correct stage yield and accuracy when present

The full leaderboard adds coverage, resolution, latency, and cost.

#ProviderCorrect stage yieldAccuracy when present
1Fiber77.2%84.2%
2Apollo56.0%70.8%
3People Data Labs49.3%65.3%
4PredictLeads43.3%84.1%
5CompanyEnrich35.8%44.0%
6Explorium14.9%36.7%
7Ocean.io4.1%11.7%

Most accurate funding data — common questions

Which company funding data provider is the most accurate?

It depends which accuracy you mean. For the most correct latest funding stage end to end — returning the right stage across all companies — Fiber leads at 77.2% correct stage yield. For the fewest wrong stages among the ones a provider actually returns, Fiber leads at 84.2% accuracy when present. Both are judged against source-verified funding events on the same 300 companies — a measured answer, not a vendor's self-claim.

What does 'accurate' mean for funding data — coverage or correctness?

Two different things, and vendors blur them. Accuracy when present asks: when a provider returns a stage, how often is it right? Correct stage yield asks: across every company with a verified stage, how often does the provider return the right one — so a provider that stays silent is penalized. A vendor can look 'accurate' by only answering easy companies; correct stage yield closes that loophole. Fiber leads on it (77.2%).

Why not just trust each vendor's stated accuracy number?

Because every vendor tests on its own sample and grades itself first, and 'accuracy' is measured differently each time. This comparison fixes both: the same 300 identity-verified companies, one canonical stage contract, every stage judged against a primary source (company newsroom, wire announcement, or filing). No vendor pays for inclusion, and inputs, outputs, and judgments are public.

How is funding-stage accuracy verified?

Each company's latest funding stage is independently researched and tied to a primary source before any provider is scored. Provider responses are normalized to one canonical stage contract (pre-seed through Series H, growth, PE, debt, grant, and other equity outcomes), then each returned stage is compared to the source-verified reference. Missing references are excluded from the denominator; wrong or missing stages lower yield.