Clay Waterfall Enrichment: A Step by Step Guide

Quick answer: what is Clay waterfall enrichment

Clay waterfall enrichment is a method for filling contact and company data by running several enrichment providers in sequence, one after another, so each source only gets asked for what the last one could not find. You start with the cheapest or most reliable source, escalate to pricier tools only for the gaps, and stop paying multiple providers to find the same email address. It is built inside Clay using tables, conditional logic and API waterfalls that pull from tools like Apollo, Clearbit and Phantombuster, ordered by cost and coverage rather than by whichever tool you happen to like most.

Most teams do the opposite by accident. They buy one enrichment tool, accept whatever it returns, and treat the gaps as a data quality problem rather than a sequencing problem. Updated August 2026.

What you need before you start

You cannot build a waterfall with nothing to pour. Get these sorted first.

• A Clay account with enough credits to run a test batch of at least 200 rows
• A defined ideal customer profile, so you are enriching the right accounts and not just any accounts
• API keys or Clay native integrations for at least two enrichment providers
• A source list with something to match on: domain, LinkedIn URL or company name
• A clean view of which fields you actually need, rather than enriching everything a provider offers

If your ICP is still a vague paragraph rather than a scored list, sort that out before you touch Clay. Our guide on ICP scoring and GTM engineering covers how to build the scoring layer that should sit underneath any enrichment work.

The steps: building your first waterfall

Step 1: map the fields you need

List every field the deal actually depends on: verified email, direct dial, company headcount, funding stage, whatever matters to your outbound. Do not enrich fields you will never use. Every extra column is a cost you are paying for curiosity, not for pipeline.

Step 2: order your providers by cost and coverage

Put the cheapest, highest hit rate source first. Send only the rows it missed to the next provider. Send only what is still missing to the third. This is the entire trick. A comparison like Apollo versus ZoomInfo is useful here, since the two behave very differently depending on region and company size.

| Tier | Typical role | Example tools | |---|---|---| | First pass | Cheap, broad coverage | Apollo, Clearbit | | Second pass | Fills gaps the first pass missed | Phantombuster, niche verticals | | Third pass | Expensive, used sparingly | Manual research, premium data APIs |

Step 3: set the waterfall logic in Clay

In Clay, this means a column that checks whether the previous enrichment returned a value. If it did, skip. If it did not, call the next provider. Clay's conditional columns handle this natively, so you are not writing custom code, just sequencing logic.

Step 4: test on a small batch

Run 100 to 200 rows before you touch the full list. Check hit rate per provider and cost per filled field, not just cost per row. A provider that fills 90% of what it is asked for on cheap credits beats one that fills 95% on expensive ones.

Step 5: watch cost per enriched row, not cost per credit

Providers price by API call, not by success. A waterfall that calls every provider on every row defeats the point. The whole design exists to stop that.

Step 6: scale once the sequence holds

Once the test batch behaves, push the full list through. Kitto used this kind of structured enrichment to map 2,400+ contacts and launch a campaign in 11 days, according to our Kitto case study.

Why bother with a waterfall instead of one provider?

Because no single provider covers everyone. Each has strengths in certain industries, regions or company sizes, and gaps everywhere else. A waterfall treats that as a fact to plan around instead of a flaw to complain about.

There is also a cost argument. Calling an expensive provider for every row when a cheap one would have found 70% of the data is money spent proving a point nobody asked you to prove. Krige Holdings cut data errors by 80% with no headcount added by tightening exactly this kind of sequencing, as covered in the Krige Holdings case study. Clean sequencing also cuts coordination time, which for Krige Holdings fell by 65%, per the same case study.

Do you need a waterfall for every list?

No. If you are enriching 50 accounts for a single outreach push, one good provider is fine. Waterfalls earn their keep on lists in the hundreds or thousands, where a percentage point of hit rate translates into real pounds, and where manual review of every gap is not realistic. For anyone weighing up tools before committing budget, our comparison of data enrichment tools for GTM is a reasonable place to start.

Common mistakes people make with Clay waterfalls

Most of these come from treating enrichment as a one off task rather than infrastructure.

• Sending every row to every provider regardless of whether earlier steps already found the data
• Enriching fields nobody on the sales team will ever look at
• Skipping the test batch and finding out the sequence is wrong at 5,000 rows instead of 100
• Never revisiting provider order as coverage changes over time
• Treating a low hit rate as a Clay problem when it is usually an ICP or list quality problem

Frontrunners cut their shortlist time from days to about one hour and lifted database activation by 3x once the underlying data process was fixed rather than patched, according to the Frontrunners case study. That is the difference between fixing the sequence and just adding another tool on top of a broken one.

How do you keep a waterfall from rotting

Providers change their coverage, their pricing and occasionally their entire product without telling you. A waterfall built in January can be quietly wrong by June. Recheck hit rates every quarter, not because it is fun, but because paying for a provider that no longer earns its place in the sequence is money you will not notice leaving.

Bright Entertainment mapped 800 or more decision makers across 6 segments and lifted response rate from 0.9% to 7.1%, which only holds up if the underlying enrichment stays accurate over time, as detailed in the Bright Entertainment case study.

FAQ

What is the difference between Clay waterfall enrichment and normal enrichment?

Normal enrichment sends every row to one provider and accepts whatever comes back. A waterfall sends each row through several providers in order, only calling the next one for rows the previous provider missed, which keeps cost per filled field lower.

Do I need to know how to code to build a waterfall in Clay?

No. Clay's conditional columns handle the sequencing without custom code. You do need to understand your data well enough to decide what order providers should run in.

How many providers should be in a waterfall?

Two or three is usually enough. Beyond that, you are often paying for marginal coverage that a better ICP or cleaner source list would have solved more cheaply.

Can Clay waterfall enrichment fix a bad list?

No. Enrichment fills gaps in data you already have. It cannot invent contacts who do not exist or fix a list built against the wrong ICP in the first place.

Getting this right without doing it yourself

Setting up a waterfall properly takes a few hours the first time and about ten minutes every time after that. Most teams either never get past the first attempt or build something that quietly wastes credits for months before anyone checks. Over 96% of pages published online get no organic search traffic at all, according to Ahrefs' study of 14 billion pages, which is a fair analogy for enrichment nobody ever audits: built once, forgotten, quietly useless.

If you would rather have this built properly than half configured on a Tuesday afternoon, that is the kind of work we do at Zuun Global. You can read more about how we approach it on our services page, or see how the founder ended up doing this for a living on the about page. If you have more questions before committing to anything, the FAQ page covers the practical ones.

Enrichment is not glamorous. It is plumbing. Good plumbing means nobody thinks about it, which is exactly the point. Get started and we will build the waterfall so you do not have to think about where your data comes from again.