Waterfall enrichment explained: how it works, with a worked example
What waterfall enrichment is, how the provider cascade runs step by step and a worked example showing match rate gains and where the cost stops being worth it.
ON THIS PAGE · 8 SECTIONS
THE SHORT VERSION
- Waterfall enrichment means querying several data providers one after another, in a fixed order, instead of trusting a single source.
- One provider alone verifies 55% to 70% of a typical B2B list. Three or four providers in sequence push that past 85%.
- Each added provider recovers less than the one before it, so most waterfalls stop at three or four sources, not twenty.
- Verification has to run on every candidate match before you send to it or you just trade a coverage problem for a bounce problem.
- It costs more per record than buying from one provider. Plan for that cost before you build the list, not after the first bounce report.
Waterfall enrichment means you send a contact record through more than one data provider, one at a time, in an order you set in advance. You call the next provider only when the one before it comes back empty or fails verification. The record keeps moving down the list until you get a confirmed email or phone number or until every provider has been tried and it still has nothing. The reason this exists is simple. No single B2B data provider covers your whole market. One source gets you part of a list. A waterfall gets you most of it, at the cost of a few extra lookups per contact.
This piece walks through why one provider is never enough, how the cascade actually runs step by step, a worked example on a 500 account list and where waterfalls still fail even when you build them correctly.
What waterfall enrichment actually is
A waterfall is not a single big database. It is a sequence. You take a record that is missing a field, usually a verified email or a direct phone number and you send it to provider one. If provider one returns a match and that match passes verification, you stop and write the result back to the record. If it does not, the record falls through to provider two, then provider three and so on until something verifies or the stack runs out.
This is different from the all in one model that large databases like ZoomInfo or Apollo sell, where one vendor tries to be every source at once. It is also different from querying every provider in parallel and picking whichever answers first, which is faster but burns a paid lookup on every provider for every record whether you need it or not. A waterfall is slower per record and cheaper per list, because you only pay for provider two when provider one actually failed.
Why one provider is never enough
Every data provider builds its database from a different mix of public filings, scraped profiles, partner feeds and user submitted contact books. That mix has gaps. A provider that is strong on US software companies can be weak on manufacturing in Germany. A provider that is strong on LinkedIn derived titles can be weak on direct dial phone numbers. None of this shows up until you run your actual list against it.
Source: Unify, 2026
Contact data also goes stale faster than most teams expect. B2B databases lose between 22.5% and 70% of their accuracy within a year, depending on the field, with email addresses decaying fastest because people change jobs, companies change domains and inboxes get retired. A single provider that was right six months ago is not automatically right today. A waterfall does not fix decay by itself, but it gives you more than one shot at a record that has gone stale in one source but not in another.
Source: ZoomInfo, 2026
How the cascade runs, step by step
- Input. A record enters the waterfall with whatever you already have, usually a full name plus a company domain or a LinkedIn profile URL.
- First pass. The record is queried against your primary provider, the one with the strongest coverage and the lowest price for your specific market and industry mix.
- Fallback. If the primary provider returns nothing or returns a match that fails verification, the record moves to the second provider. If that fails too, it moves to the third.
- Verification. Every candidate email runs through a syntax check, an MX record check on the domain and an SMTP handshake before you accept it, not after you have already sent to it.
- Exit. Verified fields get written back to the record. Anything still unresolved after the last provider gets flagged for a short manual check or dropped from the send list entirely.
The order matters more than the count. Put your most accurate and cheapest source first, since that is where most of your volume should resolve. Put a broad, lower accuracy source last, since by that point you are mopping up the records nothing else could find and a rough match is better than none.
A worked example: 500 accounts, four providers
Here is an example run, built to show the shape of the numbers rather than any real client's result. Say you have a list of 500 target contacts pulled from an account list you already built. You run them through a waterfall of four sources in priority order.
| Step | Source | New verified matches | Running total |
|---|---|---|---|
| 1 | Primary paid provider | 300 of 500 | 60% |
| 2 | Second provider | +100 | 80% |
| 3 | Third provider | +45 | 89% |
| 4 | Manual LinkedIn check | +15 | 92% |
An illustrative run. Each added source recovers less than the one before it, which is why most waterfalls stop at three or four providers instead of twenty.
That shape is not specific to this example. Adding a second provider to a waterfall typically recovers 15% to 25% of the records the first provider missed. A third provider usually recovers another 8% to 12%. A fourth adds only 3% to 5% more. Past that point you are paying for lookups that mostly come back empty, so the last 5% to 8% of a list is often cheaper to research by hand than to keep adding providers for.
Source: Unify, 2026
A waterfall does not find you more buyers. It finds you more ways to reach the buyers you already picked.
Verification is not optional
Coverage without verification is a trap. A provider can return a formatted email address that looks correct and still be wrong, because the person left the company, the mailbox was closed or the domain changed its mail server. Verification catches that before it becomes a bounce. The three checks that matter are a syntax check on the address format, an MX record lookup to confirm the domain accepts mail at all and an SMTP handshake that asks the receiving server whether the specific mailbox exists without actually sending a message.
The payoff shows up directly in your sending metrics. Single-source enrichment without this step typically produces bounce rates between 8% and 15% at scale. Teams running a verified waterfall report bounce rates dropping under 3%. That gap is the difference between a sending domain that stays healthy and one that gets throttled by mailbox providers within a few weeks.
Source: Unify, 2026
Building one without overspending
You do not need twenty providers and you do not need to build this before you know who you are enriching. Map the account list first. If you have not done that yet, our piece on how we map a market before outbound starts covers the scoring and filtering that should happen before a single record reaches a data provider. Enriching accounts you should not be targeting in the first place is the most common way teams waste a provider budget.
THE USUAL WAY
- Buy a big list from one provider and hope the coverage holds.
- Send without verification, then react to the bounce report.
- Re-buy the whole list when accuracy drops, without knowing which fields actually went stale.
THE SYSTEM WAY
- Two or three providers in a fixed order, cheapest and most accurate first.
- Verify every match before it reaches a sending tool.
- Re-run only the accounts that matter now, prioritised by the signal driving the outreach.
That last point matters more than it looks. Enrichment is not free, so run it closest to the moment you actually plan to reach out. If you are prioritising accounts by trigger events rather than working the whole list at once, our guide to B2B buying signals that actually predict a deal explains how to rank which accounts deserve a fresh enrichment pass first. Enriching an account that will not get touched for another two months just means the data decays before you ever use it.
In the system we build for clients, the enrichment waterfall sits right after account scoring and right before the first message goes out. See how that fits with the rest of the system.
FROM THE YARD
Start with two providers, not four. Measure the actual match rate on your real list before you add a third. The percentages in every vendor's blog post are averages across every industry they sell into. Your market might clear 90% on two providers. It might need five instead. You only find out by running it.
Where it breaks down
I want to be honest about the limits here, because every provider selling enrichment tools will tell you a waterfall solves coverage completely. It does not. Four problems show up consistently.
First, cost per record climbs with every added provider. Most providers charge per lookup whether or not they return a result. A waterfall with five sources can cost more per verified contact than a single premium provider, especially once you count the lookups that come back empty.
Second, speed drops. Querying three providers in sequence with verification on each step takes longer than one lookup. For a list of a few hundred accounts this does not matter. For tens of thousands run on a deadline, it does.
Third, some regions and industries stay thin no matter how many providers you stack. Private companies outside the US and UK, government adjacent organisations and small manufacturing firms are under-covered across almost every provider we have tested, because the underlying public records those providers scrape simply are not as complete there.
Fourth, a verified email is not the same as an engaged one. Verification confirms the mailbox exists. It says nothing about whether that person still opens mail, still holds that role or still cares about the problem you are solving. Treat a high match rate as permission to reach out, not as a guarantee anyone will reply.
Questions about waterfall enrichment
How many providers should a waterfall include?
Most teams get the best return from two to four providers in priority order. Each added source recovers less than the one before it, roughly 15% to 25% from a second provider, 8% to 12% from a third and only 3% to 5% from a fourth, so beyond four sources you are usually better off researching the leftover records by hand.
Does waterfall enrichment replace a TAM list?
No. Enrichment fills in contact details on accounts you have already decided to target. It does not decide who belongs on the list. Build and score your account list first, then run enrichment only on the accounts that passed that filter.
What is the difference between waterfall enrichment and an all in one database?
An all in one database like ZoomInfo or Apollo tries to be a single source for every field. A waterfall deliberately combines several smaller or specialised sources in sequence, because no single database covers every market, role and region equally well.
How often should a list be re-enriched?
B2B contact data decays at roughly 22.5% to 70% a year depending on the field, so a list that is enriched once and never touched again starts going stale within weeks. Re-run enrichment on accounts close to outreach rather than on a fixed calendar, so the data is freshest right before you need it.
Does waterfall enrichment work outside the US?
It works, but coverage drops. Most providers are strongest on US and UK companies and weaker on private firms elsewhere. Outside those markets expect to add a local or regional provider to the stack and to lean more on manual verification for the accounts that still come back empty.
Sources: Unify, waterfall enrichment and B2B data, 2026, ZoomInfo, B2B data decay rates and costs, 2026