If you have spent any time inside Clay, you already know the platform's real superpower is not scraping — it is waterfall enrichment. Yet most teams set it up wrong: they bolt on one provider, watch a third of their list come back blank, and conclude "the data is bad." The data is fine. The routing is the problem.
At GenFlows we run waterfall enrichment on every outbound list we touch for e-commerce and B2B clients. This is the exact, repeatable build we use — written for the GTM engineer who wants verified emails, not a vendor pitch. By the end you will have a working waterfall, a sane provider order, and a verification step that keeps junk out of your sending infrastructure.
A "waterfall" runs your enrichment providers in sequence rather than all at once. Clay asks provider #1 for the email. If it returns a valid result, the waterfall stops and moves to the next contact. If provider #1 comes back empty, Clay "falls" down to provider #2, then #3, and so on until something hits or the list is exhausted.
The economic difference is the whole point. Querying five providers in parallel means paying five times for every contact. A waterfall means you pay for provider #1 on the contacts it finds, and only spend on provider #2 for the contacts that failed #1. Same coverage, a fraction of the cost.
No data vendor has complete coverage. Each one is strong in different regions, company sizes, and seniority bands. ZoomInfo skews to enterprise North America; some EU-friendly providers win on GDPR-era records; mobile-focused vendors win on direct dials. Bet everything on one and you inherit that one vendor's blind spots.
| Approach | Typical verified-email coverage | Relative cost / contact |
|---|---|---|
| Single provider | 55–70% | 1× |
| All providers in parallel | 88–95% | 4–5× |
| Waterfall (ordered) | 88–95% | ~1.3–1.6× |
That last row is why waterfall exists: near-parallel coverage at near-single-provider cost.
Import or scrape your target contacts into a Clay table. At minimum you want Full Name, Company Domain, and ideally LinkedIn URL. Domain quality matters most — a missing or wrong domain is the single biggest cause of "no match." Normalize domains (strip www, http) before enriching.
Add a column → choose your cheapest accurate provider as the "Find Work Email" source. Map its inputs to your name and domain columns. Run it on a 25-row test slice first — never the full list. Note the hit rate.
Add a second "Find Work Email" column from a different provider. In its settings, set the run condition to only fire when the first column is empty (e.g. Provider 1 Email is empty). This is the mechanic that turns two columns into a waterfall — provider #2 only spends credits on the misses.
Repeat Step 3 for each additional provider, each gated on "all previous columns empty." Three to five total is the sweet spot; beyond that you are paying premium credits to chase a few percent.
Add a formula column that coalesces the results: return the first non-empty email across your provider columns. Now you have one clean field to work with regardless of which provider found it.
Add an email verification column (e.g. a verifier provider) on top of Final Email. Keep only valid / deliverable results. This single step protects your sender reputation more than any other — catch-alls and risky addresses never reach your sequencer.
Filter to verification = valid, then push to your CRM or sequencer via Clay's native integrations or a webhook. If you automate that hand-off, see our companion guide on connecting Clay to n8n.
Order by cost-to-accuracy, not reputation. A practical default we use as a starting point (then tune against your own hit rates):
| Tier | Role in the waterfall | Why here |
|---|---|---|
| 1 | Cheap, high-volume provider | Clears the easy 55–65% at lowest cost |
| 2 | Region/industry specialist | Fills gaps the generalist misses |
| 3 | Premium accuracy provider | Expensive, so only runs on the hard remainder |
| 4 | Email verifier | Quality gate before export |
Re-order after your first 500 contacts. Whichever provider has the best hit rate on your ICP earns the top slot — this is data-specific, so let your numbers decide.
Three to five. Two providers already lift coverage meaningfully; four to five is the point of diminishing returns for most B2B and e-commerce lists.
Each enrichment provider has its own credit cost, but run conditions ensure later providers only spend on earlier misses — so total spend is far lower than running providers in parallel for the same coverage.
For a clean, well-targeted list, 88–95% verified work emails is achievable. If you are under 75%, the problem is almost always domain quality or provider order, not Clay itself.
Yes — the same pattern works for mobile/direct dials. Use dial-specialist providers and gate each on the previous column being empty.
Want this built for you? GenFlows runs waterfall enrichment, verification, and sequencing as a managed outbound system. See how the major data providers compare, or talk to our team about a done-for-you build.
By the GenFlows GTM engineering team. Last updated June 2026.