The same 282 domains
We sent every domain to the EnrichLoops API on October 4, 2026 and got data back for 243 of them (86.2%). The 39 with no data still count in the success score.
Company enrichment benchmark / 01
How our company data holds up on a public benchmark
OpenBenchmarks publishes a company enrichment test with 282 domains and human-checked answers. We ran the same domains through our API, scored the output with their rubric, and lined it up against the 11 providers they already test.
We ran and scored this ourselves without OpenBenchmarks’ judge models, so it isn’t an official OpenBenchmarks result and they haven’t endorsed it.
EnrichLoops / test report
Returned-field accuracy
92.4%
Share of the fields we returned that were right,
averaged per company.
- Enrichment success
- 78.1%
- Companies with data
- 243 / 282
Estimated usage cost$4.86 USD
243 credits at the Starter rate
Starter plan is $25.00/month
October 4, 2026Scored by EnrichLoops
01 / The comparison
Where we land against the other providers
Switch between success and accuracy to re-sort the chart. The cost column is what each provider’s run of the same domains would cost at entry-level pricing.
The share of fields each provider got right, out of every field with a reference answer, averaged across all 282 companies. A blank counts as wrong.
Provider / evaluationScoreEst. cost (USD)
People Data LabsPublished
89.0%$27.70ApolloPublished
88.6%$10.88ParallelPublished
86.0%$14.10Predict LeadsPublished
80.6%$40.00EnrichLoopsOur run
78.1%$4.86ExploriumPublished
70.5%$22.00CompanyEnrichPublished
69.7%$2.76Nimble standardPublished
69.3%$1.41ZoomInfoPublished
63.9%$93.45ExaPublished
60.4%$4.23Nimble litePublished
54.0%$0.31SeltzPublished
30.0%$1.41Cost is an estimate for each provider’s run on the same 282 domains. It isn’t a subscription bill, and each provider returned a different amount of data. For EnrichLoops, 243 credits at $0.02 each come to $4.86, before free credits or tax. The Starter plan is $25.00 a month. A few published figures have caveats: Nimble’s leaves out the extraction model, Predict Leads’ includes its $40 minimum, and Apollo’s and Explorium’s are estimated from usage. More on cost
We scored the EnrichLoops run ourselves. Every other row comes from OpenBenchmarks’ published results as of October 4, 2026. The evaluators and run dates differ, so read this as context, not a place on their leaderboard. How we scored it
What our run cost
$4.86 of credits for 243 company profiles
Each profile we returned used one credit. The 39 domains we had no data for cost nothing.
- Credits charged
- 243
- Starter rate per credit
- $0.02
- Estimated usage value
- $4.86 USD
The Starter plan is $25.00 a month for 1,250 credits, which works out to $0.02 a credit. So $4.86 is the value of the credits this run used, not a bill. On Starter you’d still pay $25.00 that month, and unused plan credits reset each cycle.
Prices checked October 5, 2026, before free credits, taxes, or volume discounts. Read cost next to success and accuracy, since every provider returned a different amount of data.
02 / Field by field
Most of our misses are blanks, not wrong answers
Field success counts a blank as a miss. Accuracy only looks at the answers we gave. The gap between the two columns is mostly companies we returned nothing for.
| Field | Correct / eligible | Field success | Returned-field accuracy |
|---|---|---|---|
| HQ locationCity and country | 210 / 275 | 76.4% | 95.5%210 of 220 returned answers |
| IndustryPrimary business category | 232 / 282 | 82.3% | 95.9%232 of 242 returned answers |
| Employee bandHeadcount range | 218 / 280 | 77.9% | 90.1%218 of 242 returned answers |
| Founded yearYear established | 163 / 210 | 77.6% | 94.2%163 of 173 returned answers |
| LinkedIn URLCompany profile | 218 / 282 | 77.3% | 89.7%218 of 243 returned answers |
Eligible means the reference has an answer for that field. Companies without one are left out of that field’s numbers.
03 / Methodology
How we ran and scored the test
We ran this ourselves, so here is what each number means and where it falls short of a controlled comparison.
Scored against their reference answers
We used OpenBenchmarks’ pinned reference data and public rubric. Exact and numeric fields were checked by script. Where HQ or industry didn’t match exactly, an AI reviewer (OpenAI Codex) applied the rubric. We didn’t run OpenBenchmarks’ own judge models.
Not a head-to-head test
The other providers’ scores come from the OpenBenchmarks snapshot taken October 4, 2026. We didn’t rerun their APIs. Run dates and evaluators differ, and so do a few reference counts, so treat the chart as context rather than a controlled comparison.
The fine print
Enrichment success (78.1%) is correct fields out of every field with a reference answer, averaged per company across all 282. A blank counts as wrong. Returned-field accuracy (92.4%) is correct fields out of the fields we returned, averaged over the 243 companies we returned data for.
We also checked whether each answer belonged to the right company. We matched the wrong company 16 times. Giving those, and one unclear match, zero credit puts enrichment success at 76.3%.
Provider scores come from OpenBenchmarks company enrichment, captured October 4, 2026. Reference answers come from their human-reviewed company dataset, scored with their public rubric.
Open results
Download the raw results
Results frozen October 4, 2026. Cost estimate added October 5.
Try it yourself
Check our numbers against
your own accounts
The free plan includes 100 credits a month, enough to test 100 of your own company domains before you pay anything. No card needed.