Contents
Waterfall enrichment is a way of finding contact data by asking several providers in turn. If the first provider has no email or phone number for a contact, the request falls through to the second, then the third, until one returns a result or the list runs out. It exists because no single data provider covers every company, every region or every job title well.
It has become a standard part of the GTM engineering stack. This piece explains how it works, what published provider tests show, and how to set the order.
How does waterfall enrichment work?
A waterfall has three parts.
1/ A sequence of providers, ordered by a rule you choose, such as accuracy first or cost first.
2/ A stop condition. Usually the waterfall stops at the first valid result, so you only pay for the providers you actually needed.
3/ A verification step at the end, which checks that what was found is real. For email, that means checking deliverability. For phone numbers, it means checking the number reaches the right person.
The result is higher coverage than one provider alone, at a cost controlled by the order you set.
Why not just use one provider?
Because provider quality changes with the segment you sell into. Clay's test of 12 email finders, published on 4 May 2026, used 1,075 contacts at companies under 500 staff and more than 3,700 at larger companies. For the smaller companies, seven providers scored above 96% confidence, led by Datagma at 97.70%. For the larger companies, the leader, Wiza, scored 84.88%.
That gap is the point. Enterprise email is harder: multiple domains, subsidiaries and many staff with similar names. A provider that is excellent for small businesses may be ordinary for enterprise accounts. Clay's report says a waterfall layering several providers "can triple your email coverage" compared with relying on one tool.
What do phone data tests show?
Phone data is harder still. Clay's mobile number test, published on 3 May 2026 with call-scoring firm TitanX, collected 9,806 mobile numbers from 10 providers for about 6,000 contacts across North America, Europe and Asia Pacific. Only 23% were rated high quality: verified numbers for people with a high likelihood of answering.
No provider won everything. Forager led on accuracy, with quality scores of 46% in North America, 38% in Europe, the Middle East and Africa, and 59% in Asia Pacific. ContactOut had the widest coverage, returning 2,878 numbers, close to half of the original contacts. On cost, Bytemine came in at 5 cents per verified number in North America, and Leadmagic at 22 cents in Europe and 18 cents in Asia Pacific.
What is a good number worth?
TitanX's scoring puts its top-rated numbers at a connection rate of at least 25%, mid-rated numbers at around 4% to 10%, and unverified numbers usually under 1%.
Run that on a rep making 100 dials. On top-rated numbers, that is at least 25 conversations. On mid-rated numbers, 4 to 10. On unverified numbers, fewer than one. Same rep, same hours, a very different day. TitanX found reps can have up to five times more conversations an hour when working from providers with more verified numbers.
This is why the verification step matters as much as the waterfall. More numbers are worth little if most of them do not connect.
How should you order a waterfall?
Clay's suggested order for mobile numbers is a useful template for any data type:
1/ Start with the most accurate providers to build a base of good records.
2/ Add volume providers to widen coverage.
3/ Fill regional gaps with specialists that are strong in the markets you sell into.
4/ Finish with low-cost providers to catch what is left.
Then adjust it to your market. Clay's tests show provider strength varies by region and company size, so a waterfall built for North American enterprise accounts will be in the wrong order for European small businesses.
What should you watch out for?
Three cautions.
1/ Read vendor tests as vendor tests. Both reports above were published by Clay, which sells waterfall enrichment. The method is laid out in detail and the data is shared, which helps, but run your own sample before committing budget.
2/ Watch catch-all domains. Clay's email report notes that standard verification tools often miss catch-all and junk addresses, which can make up to half of some lead lists. A waterfall that stops at the first "found" email can stop at a bad one.
3/ Coverage is not permission. Finding a contact detail does not make it lawful to use. Marketing rules differ by country and by type of recipient, and a waterfall applies whatever rule it was given to every record. Build the rules for each market into the workflow before it runs.
A note on Peach Data
Peach Data builds go-to-market engineering systems for sales and marketing teams: the data foundation, the workflows and the agents behind outbound and inbound. We started as a verified data business, so every build starts with the data layer. peachdata.co.uk
