Waterfall enrichment means querying data providers in sequence rather than relying on one. Ask provider A for an email address; if A returns nothing, ask B; if B misses, ask C. Stop at the first hit.
Sometimes called cascade enrichment. The idea is straightforward and the reason it works is that different providers have different gaps.
Why it raises coverage
No single B2B data provider has everyone. Each builds its database differently (some from user contributions, some from public sources, some from partnerships) so their blind spots do not overlap neatly.
A provider covering 60% of your segment and another covering 55% do not cover the same 60% and 55%. Combined, they might reach 75%. Add a third and you might get to 82%.
The gains diminish sharply. Rough shape, though actual numbers vary enormously by segment:
| Providers queried | Illustrative cumulative match |
|---|---|
| One | 60% |
| Two | 75% |
| Three | 82% |
| Four | 86% |
| Five | 88% |
The first additional provider is worth a lot. The fourth is worth a few percentage points, and you are paying for every query along the way.
The cost that surprises people
Here is the part that catches teams out: you pay for the misses too.
If provider A charges per lookup regardless of result, a query that returns nothing still costs you. Run a 1,000-record list through three providers with a 60% first-hit rate:
- Provider A: 1,000 queries, 600 hits
- Provider B: 400 queries, 150 hits
- Provider C: 250 queries, 55 hits
That is 1,650 queries for 805 records. Your effective cost per usable record is roughly double the advertised per-query rate, and that is before the records that are wrong rather than missing.
The one question to ask any provider in a waterfall: do you charge for lookups that return nothing? Some do not, and that changes the arithmetic completely.
Where it genuinely pays
International segments. The clearest case. A provider strong in North America may return 30% in Germany or Brazil. Layering providers with different regional strengths is the only practical fix.
Phone numbers. Mobile match rates are lower than email across every provider, so the marginal provider adds more.
Non-obvious roles. Operational and technical titles at smaller companies are patchily covered everywhere.
High-value, low-volume targeting. Working 200 named accounts where each is worth a lot, spend heavily per record, because the cost is trivial against the deal.
Where it does not
High volume, low deal size. At 10,000 records a month with $4,000 contracts, tripling your data cost to add 20 percentage points of coverage rarely pays.
Well-covered segments. If your primary provider returns 80% on US mid-market software companies, the waterfall is buying you the hardest 10% at the highest marginal cost.
When accuracy, not coverage, is the problem. This is the common misdiagnosis. A waterfall finds more records; it does nothing about records that are found and wrong. If your bounce rate is high despite good match rates, you have a verification problem, and adding providers makes it worse by pulling in more low-confidence data.
The three traps
1. Ordering by price instead of accuracy. Putting the cheapest provider first maximises how many records come from your least reliable source. Order by accuracy in your segment, then by price, and test which is which rather than assuming.
2. No verification step at the end. A waterfall returns whatever the first provider offers. Without verification after the cascade, you are sending to addresses that were found but never confirmed.
3. Unbounded spend. A waterfall with no cap can query every provider for every record. Set a stop rule, for example, do not query providers beyond the second for leads scoring below your ICP threshold.
That third one has an obvious extension worth stating: score before you enrich. Running a full waterfall on records you will not contact is the most common way data budget disappears. What a B2B lead actually costs covers the sequencing.
Building one
Through a platform. Several tools offer waterfall enrichment natively, handling the sequencing and billing. Simplest, least controllable.
Through a workflow tool. n8n or similar, calling each provider’s API in sequence with your own logic. Cheaper per record, requires someone technical, and it is yours to maintain.
Through an enrichment platform. Purpose-built for this, with many providers available. Powerful, and the credit model takes work to predict. Clay versus native enrichment covers that trade.
Or not at all. If your primary provider covers your segment well, the correct answer may be a single source plus verification, which is simpler and cheaper than any of the above.
How to know if you need one
Run this before building anything.
Take 100 accounts in your real target segment. Run them through your current provider. Record the match rate and verify a sample of the results.
If match rate is high and accuracy is good: you do not need a waterfall.
If match rate is low: a waterfall will help. Test a second provider on the misses specifically and see how many it recovers.
If match rate is good but accuracy is poor: you have a verification problem. A waterfall makes it worse.
The full method is in how to compare data providers, and it takes an afternoon.
The short version
Waterfall enrichment queries providers in sequence until one returns a match. It raises coverage because providers have non-overlapping gaps, with sharply diminishing returns after the second or third.
The cost is steeper than it looks, because misses frequently still cost a credit. Effective cost per usable record is commonly around double the advertised rate.
It pays for international segments, phone numbers and high-value accounts. It does not pay at high volume with small deals, in well-covered segments, or when your actual problem is accuracy rather than coverage.
Order by accuracy rather than price, verify after the cascade, cap the spend, and score leads before enriching them.
Next: how to compare data providers, or what a B2B lead actually costs.
Frequently asked questions
What is waterfall enrichment?
Querying multiple data providers in sequence, stopping at the first one that returns a match. It raises coverage because different providers have different gaps, and is sometimes called cascade enrichment.
Does waterfall enrichment cost more?
Usually around double the advertised per-record rate, because lookups that return nothing often still consume a credit. Ask each provider whether misses are charged, some do not charge, which changes the maths considerably.
When should you use waterfall enrichment?
For international segments where no single provider has good coverage, for phone numbers where match rates are low everywhere, and for high-value named accounts where per-record cost is trivial against deal size.
Does waterfall enrichment improve data accuracy?
No. It finds more records, not better ones. If your bounce rate is high despite good match rates, the problem is verification, and adding providers pulls in more low-confidence data rather than fixing it.