Waterfall enrichment: what it is, when it works, and what a miss should cost you

Waterfall enrichment is a method that sends a missing field, such as a work email or phone number, to several data providers in a set order, stops when one returns a confirmed value, and moves to the next provider only when the one before it finds nothing.

It raises coverage. It does not, on its own, raise accuracy. And the billing rule behind it decides whether a long chain saves you money or quietly multiplies the bill.

Key takeaways

• A waterfall checks providers one after another and stops at the first confirmed value. Its one guaranteed gain is coverage.
• Single sources hit a ceiling because no database holds everyone and people move. Lusha measured 12.25% of US sales leaders changing roles in twelve months.
• A waterfall cannot fix accuracy, status consistency or provenance by itself. Record the source of every field or you cannot judge any provider.
• Billing rules differ. Some vendors charge nothing when a source finds nothing. Others charge every source that returns data until a validated result appears.
• Short chains win. Each added provider returns less net new coverage, so a few vetted sources behind a large first party database is the design that holds up.

What is waterfall enrichment?

A waterfall is a fallback chain, not a single lookup. A record goes to provider A. If A returns a confident value, the run ends. If not, the record falls to provider B, then C, until something lands or the chain runs out. Query in order, stop early.

The method works on any field with more than one supplier behind it. In practice, teams use it for work emails and mobile numbers, because those fields block outreach when they are empty. It is a technique, not a product. You can build one with a few API keys, run one inside a workflow tool, or use one built into your data provider.

Waterfall enrichment is one way to do data enrichment, the broader job of filling the fields a record is missing. The difference is the number of sources. Standard enrichment asks one database and accepts the gaps. A waterfall asks several and fills more of them.

Why does single source coverage have a ceiling?

Two forces cap any single provider: what it never had, and what has changed since it last checked.

Coverage gaps are structural. Every provider builds its database from a different mix of sources, so no two hold the same people. Published industry estimates put the best single providers at 55% to 70% of a typical B2B contact list. Providers also trade quality for reach: in one 2025 benchmark, the strictest email finder scored 97% on quality and 53% on coverage, while a broader one scored 96% and 90%. You pick a side with every provider you choose.

People move. Lusha measured how fast its own records go stale, twice, against 148,000 records. In the latest run, September 12, 2026, 12.25% of US sales leaders at VP and C suite level changed roles over twelve months, and 25.67% over twenty four. The first run, on August 15, 2026, found 12.6% and 25.7%. That is about 1% a month. Full method in B2B data decay, measured twice.

You will see a higher figure, around 22% a year, in other waterfall posts. The two numbers measure different things. Lusha’s figure is a role change rate, counted in a defined cohort and published with its method. The 22% figure is an industry estimate of all decay, which counts any field going wrong, and it is rarely tied to a named study. Role change is narrower, and it is the part a waterfall can act on, because a changed role leaves a missing or stale email and phone behind.

The planning point is simple. Even a strong database misses some contacts on day one and loses about 1% of senior sales contacts to role changes each month. A second source is the fix for the first gap. A refresh cadence is the fix for the second.

How does a waterfall run, step by step?

Every waterfall follows the same path:

Input → first source → verify → cascade on miss → stop on match → flag the rest

1. Input. You pass an identity: a name plus a company domain, a LinkedIn URL, or a record ID.
2. First source. The first provider runs on every record. Its quality sets the floor for the whole chain.
3. Verify. The candidate email or phone is checked before it counts.
4. Cascade on miss. No confirmed value means the record moves to the next provider in the order.
5. Stop on match. The first confirmed value ends the run for that field.
6. Flag the rest. Records that no provider can fill are marked unresolved, not filled with a guess.

Take one masked example. J.K. is a VP of Sales with a record but no work email. The first source finds nothing. The second returns j.k@[company].com and it passes verification. The run stops. The third source never sees J.K. A second contact, R.S., has an email but no mobile. Every source misses, so the phone field stays flagged and R.S. goes into an email sequence instead of a call list. Had a source matched, the record would show a number like +1 61• ••• •••• with its source beside it.

The stop on first match rule is where the money sits. If the chain stops at the first confirmed value, you pay at most one provider per field. Not every vendor stops there. Some run until a validated result is found or every source has been tried, charge each source that returns data along the way, and charge a separate validator after each one. Both designs aim at a validated result. Only one of them can bill more than once for the same contact.

What can a waterfall not fix on its own?

Coverage is the only thing a waterfall guarantees. Three things it cannot fix by itself:

Accuracy. A waterfall stops at the first answer it gets, and it does not check whether that answer is right. Some providers return a likely pattern and call it valid. Stricter ones run more checks. Inside a chain, the forgiving providers claim a bigger share of the list, because the records they “find” never reach the stricter tools behind them. You meet those guesses later as bounces.

Consistency of statuses. Every provider labels results its own way. Accept all domains, which take mail for any address, are the classic case. One provider calls them risky, another valid with a footnote. Merge those outputs into one CRM column and “verified” stops meaning one thing.

Provenance. A waterfall is as compliant as its least transparent provider. GDPR expects you to say where personal data came from. With one provider that is one answer. With five it is five answers, and one opaque source leaves its share of records unaccounted for.

The measurement trap. The first provider sees the whole list and resolves the easy records. Later providers only see the leftovers. Provider A finds 65% of what it is asked and provider C finds 12%, but swap their places and C might find 72%. Position in the queue decides the stats, so find rate tells you little about provider quality.

The fix: record the source per field. Store which provider supplied every email and phone, then judge each provider by bounce rate and connect rate on its own records. Run one verifier at the end so “valid” means one thing. Set your accept all policy once. Start with the strictest source, even if it costs more per lookup. A vendor that marks the source on every result makes all of this possible without extra work.

Sequential, parallel, or aggregator of aggregators?

There are five ways to structure enrichment. The match rates below are published industry figures, not Lusha measurements.

The choice comes down to who answers first. Behind a large first party database, the first and cheapest answer comes from data the vendor collects and verifies itself. In a managed waterfall with no database of its own, every answer is a third party answer, and every answer carries a third party’s terms.

What should a miss cost you?

This is the question that sets the economics of a waterfall. Here is what each vendor publishes on its own pages. Full links in Sources.

Why pay on match changes the economics. When a miss costs nothing, adding a provider only costs you when that provider wins a record. The chain’s price is the number of fields found times the rate, whatever the order and however many sources run. Under Lusha’s pay on match waterfall, filling 200 missing work emails costs 400 waterfall credits whether one provider answered or the request passed through every enabled one. When a vendor charges each source that returns data, or charges a validator after each source, a long chain can bill several times for one contact before a validated result lands.

Watch the edges too. Catch all emails, the ones a mail server accepts without confirming the mailbox exists, are charged by some vendors and not others, and at least one vendor’s pricing page and FAQ disagree on the rule. Ask which rule applies before you run a large list.

How many providers is enough?

Fewer than you think. Each provider you add returns less net new coverage than the one before it, because databases overlap.

Published figures show the curve. A second provider typically recovers 15% to 25% of the records the first one missed. A third recovers another 8% to 12%. A fourth adds only 3% to 5%, at which point the cost per extra match often exceeds the value of those records. The reason is that many providers license the same upstream data, so the fourth provider in a queue often searches a near copy of a database you already tried. Coverage climbs fast on the first two levels, then flattens while spend keeps rising.

The overlap curve points to one design. Put a large first party database with published accuracy at the top, where it resolves the bulk of records. Put a short, vetted chain behind it to fill what is left. Every provider in that chain should earn its place on records the first source missed, and should sit under the same terms and compliance rules as the first source. A chain of twenty unvetted sources adds little coverage after the third and adds a provenance question for every one.

Build, buy, or built in?

You have three ways to run a waterfall.

Build it yourself. Wire provider APIs together in n8n, Make or a workflow tool. You choose every provider and the order. You also own the upkeep: API keys, routing logic, error handling, rate limits, status normalization and source tracking. Published estimates put that at four to eight hours a month for a three provider chain, and the hidden maintenance pushes the break even far higher than the sticker price suggests.

Buy a packaged waterfall. Several vendors sell the cascade as the product. You upload a list or call an API and the vendor runs the chain across third parties. Convenience is priced in, and every answer comes from a third party.

Use one built into your data provider. The waterfall runs behind the provider’s own data, so a missing field falls through to partners inside the same account, contract and credit balance. This is how Lusha’s Data Waterfall works.

When DIY makes sense: you run outreach at agency scale, you already pay for several providers, you need an order tuned to a niche market, and someone on the team owns the setup every week. If none of those apply, a built in waterfall removes the maintenance tax.

How do you evaluate any waterfall?

Ask these ten questions of any vendor. Lusha answers yes to all ten, and the next section shows how.

1. Does it check its own data first? A first party source with published accuracy should answer before any third party.
2. Does it stop on the first confirmed match? If not, find out how many sources can bill for one contact.
3. Does it charge on a miss? Get the billing rule in writing, including validator and catch all charges.
4. Can you see the source for every field? Without it, you cannot judge providers or answer a data subject request.
5. Can you opt a single call out? Some requests need first party data only.
6. Who orders the providers? A vendor set order removes tuning work. A user set order gives control and adds upkeep.
7. Can you disable a provider? You should be able to switch any one off at any time.
8. Are providers under the vendor’s terms and DNC rules? Every source should sit under the same contract, privacy terms and Do Not Call checks.
9. What accuracy does the first source publish? The first source fills the bulk of the list, so its accuracy is your accuracy.
10. Does it run the same way across UI, API and agents? A waterfall that only runs in one surface leaves gaps in the others.

How does Lusha’s waterfall work?

Lusha’s Data Waterfall completes contacts Lusha already holds. Here is how it runs, from the Data Waterfall docs and the Waterfall Enrichment page.

Lusha data first. Every request starts with Lusha’s own 515M+ verified contacts and 29M+ companies, with 98% email accuracy and 85% US / 86% global phone accuracy, and Do Not Call flags on every phone.

Then a vetted set of providers. When a work email or phone is still missing, the waterfall checks additional providers in an order Lusha sets to favor data quality. Each provider sits under Lusha’s Supplementary Terms. Admins see the full list under Account > Waterfall and can turn each provider on or off at any time.

Stop on the first confirmed match. The run ends at the first successful match for that field.

Pay only on a match. A match from an additional provider costs 2 credits per work email and 8 per phone. A provider that returns nothing costs 0 credits. Lusha’s own data is billed at standard rates, listed on the pricing page.

Control per call. The waterfall runs automatically on every eligible enrich request once an admin turns it on. Pass waterfallEnabled: false to skip it for a single API call.

Source on every result. Each email and phone carries a source value, Lusha or the provider. In Workspace, the source is marked on the result and the credit cost shows before you confirm.

Same waterfall everywhere. It runs in Lusha Workspace, through the API, and through the Lusha MCP server for Claude and ChatGPT, on the same account setting and credit balance.

Completes, does not find. The waterfall fills missing work emails and phones on contacts Lusha already has. It does not search for new contacts. Pair it with Partial Profiles in the Search API to surface contacts Lusha holds without contact details, then let the waterfall fill them.

Plans. Data Waterfall is available on Pro plans and above.

What it has filled so far. In September 2026, the waterfall filled 1 in 2 missing work emails across all waterfall accounts. On one 1,103 contact customer list, email coverage went from 90% to 97% before and after the waterfall.

FAQ

What is waterfall enrichment?

Waterfall enrichment sends a missing field, such as a work email or phone, to several data providers in a set order. The first provider that returns a confirmed value ends the run. If a provider finds nothing, the next one is tried. The goal is coverage: several sources together fill more records than any one of them. It is a method, not a product, so you can build it, buy it, or use one built into your data provider.

Does waterfall enrichment improve accuracy or only coverage?

Coverage. A waterfall stops at the first answer it gets and does not check whether that answer is right, so providers with looser standards can fill more of your list. Accuracy comes from the first source and from verification. Put a database with published accuracy first, run one verifier on every result, record the source of each field, and judge providers by bounce rate rather than find rate.

What does a failed lookup cost?

It depends on the vendor. Lusha charges nothing when a provider finds nothing, and 2 credits per work email or 8 per phone only on a match. Some vendors charge each source that returns data until a validated result appears, and charge a connected validator separately. Ask for the rule in writing, including catch all charges.

How many providers should a waterfall have?

Two to four. Published figures show a second provider recovers 15% to 25% of the first provider’s misses, a third 8% to 12%, and a fourth only 3% to 5%, because many providers license the same upstream data. Behind a large first party database, a short vetted chain fills the gaps without adding a provenance question for every extra source.

Is waterfall enrichment GDPR compliant?

It can be, but a waterfall is only as compliant as its least transparent provider. GDPR expects you to say where personal data came from and on what basis you process it. Choose a waterfall whose providers all sit under the vendor’s own terms, record the source of every email and phone, and check each provider’s methodology. Lusha marks the source on every waterfall result and runs every provider under its Supplementary Terms.

What is the difference between waterfall enrichment and a waterfall of waterfalls?

A waterfall of waterfalls is an informal name for a service with no database of its own that cascades across many third party providers per contact. A standard waterfall behind a data provider starts with that provider’s own verified data and uses third parties only to fill a missing field. Lusha’s Data Waterfall is the second kind.

Can AI agents run waterfall enrichment?

Yes. Lusha’s waterfall runs through the Lusha MCP server for Claude and ChatGPT, on the same account setting and credit balance as the API, including providers that answer in the background. Other vendors expose their waterfalls through their own MCP servers or agent connectors, so check which surfaces the waterfall covers before you rely on it.

How is Lusha’s waterfall different from most others?

Three things. Lusha’s own verified data answers first, so the bulk of records never touch a third party. The run stops at the first confirmed match. And a match costs 2 credits per work email or 8 per phone while a miss costs nothing, so adding a provider never adds a bill unless it adds a result.

Sources

• Lusha, Data Waterfall (Waterfall Reveal)
• Lusha, API FAQ
• Lusha, Waterfall Enrichment
• Lusha, B2B data decay, measured twice
• Lusha, B2B data decay rate: what we measured against 148,000 records
• Apollo Help Center, Waterfall Enrichment Overview
• Clay, The complete guide to waterfall enrichment and Clay FAQ on credits
• FullEnrich, Pricing and Help Center, How do credits work
• Explorium, Waterfall Enrichment Explained
• Unify, What Is Waterfall Enrichment?
• Hunter, Waterfall Enrichment: What It Is, How It Works, and When It’s Worth It
• BetterContact, Pricing and FAQ