What Is Waterfall Enrichment? Definition & How It Works
Waterfall enrichment explained: querying multiple data providers in sequence to boost match rates, with a worked example, trade-offs, and setup tips.
Inhalt
You enrich a list of 500 contacts through one provider and get emails back for 280 of them. The other 220 come back empty — not because the people don't exist, but because that one provider simply doesn't have them in its database. This is the exact problem waterfall enrichment was built to solve, and it's become the default setup for teams who refuse to accept a single vendor's coverage gaps.
If you've seen the term in a Clay workflow or a RevOps Slack channel and want the short version before you build one, this guide covers it: what waterfall enrichment means, how the sequence actually runs, a worked example with real numbers, and where it fits next to a real-time data enrichment tool like Enrich-CRM.
What is waterfall enrichment?
Waterfall enrichment is the practice of querying multiple data providers in sequence for the same record, stopping as soon as one returns a match. Provider A is tried first; any record it can't fill falls through to provider B, then provider C, until the field is populated or every source is exhausted. The goal is a higher overall match rate than any single provider delivers alone.
How a waterfall enrichment sequence works
The mechanics are always the same three-step loop, repeated per field or per record:
- Query the first provider. The record — usually an email, domain, or name plus company — is sent to whichever source is cheapest or historically strongest for that field.
- Check the result. If the provider returns a verified value, the sequence stops there. No further calls are made, which is what keeps costs down.
- Fall through on a miss. If the field comes back empty or unverified, the same record moves to the next provider in the stack, and the check repeats.
Teams typically order providers by a mix of cost and coverage: the cheapest or most reliable source goes first, pricier specialist sources sit further down the waterfall and only get called for the records everyone else missed.
A worked example
Say you're enriching 1,000 contact records for direct phone numbers, and you've stacked three sources by cost:
| Step | Provider | New matches | Cumulative match rate |
|---|---|---|---|
| 1 | Primary database (lowest cost) | 520 | 52% |
| 2 | Secondary database | 210 | 73% |
| 3 | Specialist / premium source | 90 | 82% |
No single source in this stack cleared 60%. Chained together, the same 1,000 records finish at 82% — the entire value of waterfalling in one row. The trade-off is visible too: step 3 is usually the most expensive call per record, and it's only worth making because it's applied to the 270 records the first two steps already failed on, not to all 1,000.
Why teams build waterfalls instead of picking one provider
Every static database is built by crawling and buying data at different points in time and from different sources, so no two providers cover the same companies or contacts equally well. One might be strong on European SMBs and weak on US enterprise; another the reverse. Waterfalling doesn't fix the freshness problem of static data — a record that comes back from step 3 can still be months out of date — but it does directly attack the coverage problem: more of your list ends up filled in, period.
This is also why the pattern is so tightly associated with tools like Clay, which was built around letting users chain several enrichment providers — Enrich-CRM among them — into a single table and configure exactly this kind of fallback logic per column.
The trade-offs
Waterfall enrichment isn't free, in either sense of the word:
- Cost stacks per fallback step. Each additional provider in the chain adds its own per-record price for the fraction of records that reach it. Badly ordered waterfalls — expensive sources called first — burn budget fast.
- Setup and maintenance take real effort. Someone has to configure the sequence, monitor which providers are actually earning their place in the stack, and re-order it as pricing or coverage changes.
- It still inherits static-data staleness. Waterfalling improves how much data you get back, not how current that data is. A contact who changed jobs last month can still get matched — correctly, to their old employer — by step 2.
- Match rate isn't the same as accuracy. A higher fill rate feels like a win, but it's worth spot-checking a sample from each step, since sources further down a waterfall are often there precisely because they're the long tail of coverage, not always the most reliable.
Waterfall enrichment vs. a single real-time source
The alternative to stacking multiple static databases is querying one source that searches the live web at request time instead of serving a stored snapshot — which is the model Enrich-CRM runs on for its 250+ company and 50+ contact datapoints. It doesn't eliminate the case for waterfalling: plenty of teams keep Enrich-CRM as one step in a Clay waterfall precisely to combine broad live coverage with the specialist sources they already pay for elsewhere. But it does change the calculation on the freshness side — a match returned live doesn't carry the "current as of whenever this database was last crawled" caveat that every static step in a waterfall does.
FAQ
Is waterfall enrichment the same as data enrichment?
No — data enrichment is the general practice of appending missing attributes to a record. Waterfall enrichment is one specific technique for doing it: chaining several providers in sequence to raise the match rate, rather than relying on a single source.
How many providers should be in a waterfall?
There's no fixed number — most working setups use two to four. Beyond that, each additional step tends to add cost and maintenance faster than it adds new matches, since the records reaching step 5 are usually the hardest ones left.
Does waterfall enrichment cost more than using one provider?
Per record, yes, for any record that needs to fall through more than one step — you're paying for each attempt. In practice teams manage this by ordering the cheapest source first and reserving pricier specialist providers for the tail end of the list, so the extra cost only applies to the harder-to-match fraction of records.
Can I combine a waterfall with a real-time source?
Yes — this is a common setup in tools like Clay, where a real-time source such as Enrich-CRM sits as one provider in the sequence alongside static databases, giving you both breadth of coverage and a source that reflects current data rather than a stored snapshot.
Comparing how a waterfall's match rate stacks up against a single B2B data provider is easiest with your own list. Create a free Enrich-CRM account — 100 credits per month, no credit card required.