Set up the column
In your table: Add enrichment → HTTP API. Then, for every recipe:- Method and Endpoint as given below. Type
/in the endpoint to insert a column’s value. - Headers: key
X-API-Key, value your Prefetch API key. - Run on a few rows, then choose fields to return: Clay lists the response’s fields, and each one you pick becomes its own column.
- Rate limit: keep requests in flight within your plan’s concurrency — on Free, 1 at a time (30 a minute); on Starter, up to 10 at a time. Rows that hit the limit fail with
CONCURRENCY_LIMITEDand can simply be re-run; failures are never charged.
Company name → domain
Most lead lists start with names./resolve turns them into domains, which every other recipe needs.
Treat a confidence of 0.8 or above as the answer. Add a Clay condition to only run the next columns when it is, and review the rest. 1 credit per row from the index, 3 when the live lookup runs.
Domain → full company profile
One/enrich call gives brand, company and industry together — cheaper than three separate calls.
12 credits per row. Well-known companies are answered from curated data in milliseconds; others take 5–20 seconds while the site is read.
Industry only, for ICP scoring
When all you need is the segment,/classify costs less than half of /enrich.
Map
data.classification.category (IAB category, e.g. Business and Finance), data.classification.subcategory and data.classification.iab_id. 5 credits per row.
Logos without an API call
For a logo to look at, you don’t need a column that calls the API at all. Get your publishable token once:Thousands of rows
Clay calls the API once per row, so a large table is many requests at your plan’s concurrency. When you only need the data once, an alternative is one batch: export the domains, send them toPOST /batch/enrich, and import the results from GET /batch/enrich/{id}. Each domain costs the same 12 credits either way.