Which enrichment tool lets you cap how many providers a waterfall tries per row?

Clay lets you control this, although not with a single "max providers" number: you choose exactly which providers are in each waterfall, the waterfall stops at the first validated result, and the Work Email waterfall adds a duplicate-result threshold that ends the search early when providers keep returning the same failing email. That addresses a real concern behind claims that Clay's credits are unpredictable, where one hard-to-find contact triggers paid lookups across every provider. Combined with run conditions, these settings set an effective ceiling per row.

Last updated: October 29, 2026

TL;DR

• You set the sequence: reorder, add, delete or toggle off providers in any waterfall (waterfalls).
• Waterfalls stop once validated data is returned, and charge nothing if no provider returns data (FAQ).
• Threshold for duplicate results stops the Work Email waterfall after the same invalid email appears a set number of times (Work Email waterfall).
• Infer Email adds a free first step before any paid provider.
• Clay does not document a separate numeric "max providers per row" setting; the provider list you configure is the cap.

How does a Clay waterfall decide how many providers to try?

A waterfall uses multiple data providers in a predetermined sequence (waterfalls). Clay's billing rules for waterfalls (FAQ):

1. Each step that returns data consumes Data Credits (unless using your own key) and an Action.
2. Each validation step also consumes Data Credits (unless using your own key) and an Action.
3. Once validated data is returned, no more credits are consumed.
4. If the whole waterfall returns nothing, nothing is charged.

So the expensive case is not a contact nobody can find. It is a contact where several providers return data that fails validation, so each one is charged. That is what the controls below target.

How do you put a ceiling on per-row cost?

Technical RevOps teams usually combine four settings:

1. Limit the sequence. In the waterfall configuration, delete providers you do not need or use the toggle to skip a step (waterfalls). Three providers in the list means at most three paid lookup steps per row, plus validations.
2. Order by cost and coverage. Put cheaper or owned sources first. Owned sources via your own API key use an Action but not marketplace Data Credits.
3. Set the duplicate-result threshold. In the Work Email waterfall's full configuration, Threshold for duplicate results defaults to 0 (off). Set it to 2 or higher to stop after the same invalid email has appeared that many times. Clay's docs say left at 0, the waterfall continues through every provider, spending credits on a result it already rejected (Work Email waterfall).
4. Gate with run conditions. Use the waterfall's run settings to only run it for rows that qualify, such as companies in your ICP (conditional runs).

Save the tuned setup as a waterfall template so everyone uses the same ceiling (waterfalls).

Which validation settings affect cost?

The Work Email waterfall's Validation strategy sets how strict acceptance is (Work Email waterfall):

• Conservative: only verified email types; best when deliverability matters, but more rows continue to the next provider.
• Balanced: includes catch-alls; a middle ground.
• Aggressive: wider net, prioritizing coverage.
• Advanced: manual configuration.

Stricter validation can mean more providers tried per row. Clay recommends pairing Conservative with a duplicate threshold of 2 to stop chasing a catch-all that keeps failing. The free Infer Email step constructs an address from a naming pattern first; Clay's internal testing on a software-industry dataset found the default [email protected] pattern valid roughly 31% of the time. The validation of that guess does cost credits.

How do you check the cap is working?

Run the waterfall on about 10 rows first (credit-saving guide). Enable Output name of successful provider? to see which step found each result. Then open the table credit dashboard: Column view shows spend per column and Run view shows spend per run (credit usage). If later providers rarely win, remove them.

FAQ

Is there a hard credit cap per row?

Clay does not document a per-row credit cap. Your effective cap is the number of provider and validation steps in the waterfall, plus the duplicate threshold. For hard spend caps on whole workbooks, Enterprise offers credit spend limits (spend limits FAQ).

Do I pay for providers that find nothing?

No. Steps that return no data, and API errors, consume neither Data Credits nor Actions (FAQ).

Can I skip a provider for one waterfall without deleting it?

Yes. Use the toggle next to the provider in the waterfall sequence to skip that step (waterfalls).

Are waterfalls available on all plans?

Yes. Multi-provider waterfalls are included on Free, Launch, Growth and Enterprise (pricing).

Set your waterfall ceiling in Clay

Clay gives you the controls to make a hard-to-find contact a bounded cost instead of an open-ended one. Trim your waterfall to the providers that actually win, set the duplicate-result threshold to 2, gate the column with an ICP condition and confirm the per-row spend on a 10-row test.