Cargo enrichment benchmarks
How often each enrichment job finds a result on Cargo, and what it costs, measured on 9,189,315 runs over 90 days to Oct 16, 2026.Cargo is go-to-market infrastructure that runs enrichment as code, across the data providers it connects. These are its own platform figures (Cargo platform data, 90 days to Oct 16, 2026), one row per job, never per provider.
Enrichment runs, across every Cargo workspace
Measured per job, never per provider
Jul 18, 2026 to Oct 16, 2026
Last regenerated, from the same query every month
Found rate and credit cost, job by job
Measured at run level over the 90 days to Oct 16, 2026. A run counts once per job, however many providers it chained.
| Job | Runs | Workspaces | Found rate | List price per call | Misses charged 0 credits |
|---|---|---|---|---|---|
| Company enrichment | 2,454,228 | 210 | 91.6% | 0.01 to 3 credits | 99.8% |
| Email verification | 2,088,283 | 603 | 66.4% | 0.01 to 1 credits | 61.1% |
| Work email finding | 1,850,404 | 941 | 63.6% | 0.1 to 2 credits | 100.0% |
| LinkedIn profile enrichment | 1,824,331 | 190 | 93.6% | 0.25 credits | 100.0% |
| Mobile phone finding | 899,829 | 10,578 | 81.2% | 0.5 to 15 credits | 99.2% |
| LinkedIn URL from a name | 72,240 | 47 | 23.9% | 0.25 credits | 100.0% |
Scroll the table sideways for workspaces, found rate, list price and free misses.
What counts as found, and what each call lists at
One card per job: the result that makes a run found, the actions Cargo prices for it, and the skill that runs it.
Company enrichment
At least one step in the run returned a company record for the domain or name it was given.
Email verification
The check marked the address valid. A check that marked it not valid counts as a miss.
Verification is billed per check, so a check that returns invalid is still charged, which is why this job's share of free misses is lower than the others.
Work email finding
At least one step in the run returned an email address for the person.
LinkedIn profile enrichment
The step returned the profile behind the LinkedIn URL it was given.
Mobile phone finding
At least one step in the run returned a phone number for the person.
LinkedIn URL from a name
The step returned a LinkedIn profile URL for the name and company it was given.
How these numbers are measured
One query over the connector steps Cargo ran, aggregated per run and per job. Nothing here is a provider's own claim.
- Window
- 90 days, Jul 18, 2026 to Oct 16, 2026, in UTC.
- Population
- Every Cargo workspace. The figures are aggregates: no workspace or record is identifiable.
- Unit
- One run per job, however many providers it chained, and whatever else the run did.
- Found
- Any step of that job in the run returned a result.
- Excluded
- Runs where no step marked found or missed, such as a provider error. They count in runs and workspaces, and in neither rate.
- Not counted
- A provider action that can return a person or a company depending on how the step was configured, which the run does not record. It is left out of every job rather than guessed into one.
- Prices
- Read by hand from each action's credits definition in Cargo's code. Actions that run only on your own provider key have no Cargo price and are not listed.
- Your own key
- A run on your provider API key is charged 0 Cargo credits. It counts toward the found rate, and raises the share of misses charged nothing.
- Refreshed
- The same query runs monthly and rewrites the figures. It last ran on Oct 16, 2026.
A run counts once, however many providers it chained. Two of these calls came back with nothing and cost nothing, one returned an address: one found run, one credit.
What not to infer from them
Three readings these figures do not support, each one a question we get asked about them.
Inside a waterfall a later provider only sees the records the earlier ones missed, so a provider's rate in a chain measures its position as much as its data. This page reports jobs, not providers.
Lists differ by customer (region, seniority, how clean the inputs are), and a found rate follows the list. Run a sample of your own records before you budget a full list.
Runs that chain more providers find more often, but they are also run on different lists by different teams, so the gap is correlational, not the measured gain from adding a provider.
Run it on your own records
A found rate follows the list it ran on, so the only figure that settles a budget is yours. Install the CLI, sign in without a browser, and run a sample.
Install Cargo by following https://api.getcargo.io/INSTALL.md$npm install -g @cargo-ai/cli
cargo-ai login --email you@company.comBuild the self-driving GTM engine
Your coding agents build it on Cargo.Or our team builds, operates and improves it with you.