In this post:
Waterfall enrichment is a method that passes each record through several data providers in a set sequence, moving to the next provider only when the previous one fails to return good data. It raises coverage by combining the strengths of many sources instead of living with one provider's gaps.
You notice the need the first time you push a list through a single provider. Even the best one comes back with a chunk of records missing, wrong, or incomplete, because no provider knows everything about every company.
Providers have different strengths, and we see this constantly in client work. One is great at emails, another at phone numbers, another at a certain region, so cascading through them captures far more than any one could alone.
TL;DR
Each record goes to your first provider and only falls through to the next when that provider comes back empty. It works like a relay rather than a vote, so no provider is asked about a record another one already resolved.
No single provider covers your whole list. Most of the extra coverage comes from the second and third sources, and the size of the lift depends on your market, so run a sample of your own list through both setups and compare fill rates field by field.
Work emails are the classic use, but direct dials and company data are just as patchy, so every important field gets its own cascade and its own provider order.
Spend stays low because the cascade stops at the first good answer. Clay refunds credits when a provider finds no email, and putting cheap or high-hit sources first means the pricier ones only see the hard records.
Choose providers whose coverage overlaps as little as possible, and treat the order, stop rules and branching (the enrichment waterfall logic) as the real work. Platforms like Clay remove the need for a contract with every source, though not the need to think.
Contact data keeps moving (median employee tenure in the US was 4.1 years in January 2026), so re-run the waterfall on a schedule and verify what it finds. The usual mistakes are poor ordering, too many providers, no verification, and running it once.
Nebor builds and tunes these cascades for clients in their own Clay workspace, field by field, and a short call is usually enough to see whether yours is leaving coverage on the table.
Following a work email down a waterfall enrichment cascade
Instead of relying on one source, you set up a cascade of them for each piece of data you need. Say you're missing a work email for a contact:
The waterfall asks your first provider. If it returns a verified email, the record is done and you move on.
If the first comes up empty, the record falls to the second provider.
It keeps falling through to the third provider and beyond, each getting a chance to find what the others missed, until the email is found or every source has been tried.
Whatever was found gets validated before it goes into your CRM or sequencer.
It's called a waterfall because records cascade down, dropping to the next source only when the one above fails. By the end, a single field like an email may have been attempted by several providers, which raises the odds you got it.
The mental model is a relay instead of a vote. You aren't asking every provider at once and comparing, you're handing the record down the line until someone returns a good answer, which keeps it both thorough and efficient.
No single provider covers your whole list
The problem is structural, so picking a better tool never solves it. Each provider gathers data in its own way, which makes it strong in some industries, regions, and data types and weak in others, with real gaps everywhere. Run any real list through one source and a share of records comes back empty or wrong.
Those gaps cost you pipeline every week. Every missing email or phone number is a contact you can't reach, so single-source coverage puts a hard cap on how much of your list you can use. The issue is incompleteness, not bad providers, and the only way past one provider's blind spots is to add others that see differently.
How much a waterfall adds depends on your list, your market, and which providers you chain, but the pattern is consistent. Most of the lift comes from the second and third providers, which pick up records the first one missed. After that, each extra source finds fewer new records while still costing money.
Aspect | Single provider | Waterfall |
|---|---|---|
Coverage | Capped at what one database holds | Each extra provider fills some of the gaps the others left |
Blind spots | Every weak region, industry, or field stays empty | Sources with different strengths cover for each other |
Cost control | One subscription, used the same way for every record | Cheap or high-hit providers go first, pricier ones only see the hard records |
Setup | One integration, little configuration | Needs a provider order, stop rules, and validation for each field |
Maintenance | Re-run one source as data decays | Re-run the cascade and review the provider order as results change |
The reliable way to size the gain is to run a sample of your own list through your current provider, then through the full waterfall, and compare fill rates field by field. It's usually the difference between reaching most of your target list and reaching only part of it.
Not just emails: a cascade for every field
People tend to link waterfall enrichment with finding emails, but the cascade is a general technique, and it pays off wherever providers struggle:
Work emails are the most common use because they're hard to find and verify, so multiple attempts clearly help.
Phone numbers make an especially strong case. Direct dials are notoriously patchy across providers, so a waterfall improves how many contacts you can call.
Company data benefits just the same. Firmographics, headcount, and technologies all improve with cascading sources, because no provider has complete company data either.
Because sources specialize, the best order differs per field: some excel at emails, others at phone numbers, others at firmographics. Clay's waterfall enrichment page reflects this, with separate provider lists for work emails, personal emails, and mobile numbers. Anywhere a single provider leaves gaps, which is everywhere, a cascade of sources fills more of them.
Where the money goes
A less obvious benefit of the waterfall is that it saves money while it improves coverage. Because the cascade stops as soon as a provider returns good data, you don't query every source for every record, which avoids paying several providers for the same field.
Clay's email waterfall is a concrete example. If a provider finds a valid email, the process stops for that contact, and if a provider finds nothing, its credits are refunded and the record moves to the next one. Not every provider bills empty lookups that way, so check before you settle on an order.
Order matters for cost as much as coverage. Putting a cheaper or higher-hit provider first means most records get resolved early, so you only fall through to pricier sources for the harder ones. A well-ordered waterfall maximizes coverage and minimizes spend at the same time, while a careless one, with an expensive or low-hit provider at the top, does neither.
Picking providers that fail on different records
Waterfall enrichment and enrichment waterfall logic sound interchangeable, and they do describe the same system, at two levels. Waterfall enrichment is the method: cascading through providers to raise coverage. The logic is the configuration that decides how the cascade behaves, covering order, stopping conditions, and branching. Deciding to run a waterfall is easy. The logic is where the skill lives, because a poorly configured waterfall wastes money.
Provider choice sits at the center of that logic. The goal we aim for is coverage that overlaps as little as possible. Two providers with identical strengths add little, while providers strong in different regions, industries, or data types each fill real gaps. Choose sources that disagree in useful ways, not similar ones that all fail on the same records.
Your own records belong in the mix too. Weighing first-party versus third-party data pays off, because combining what you already hold with external providers often produces the most accurate, complete records.
Do you need a contract with every data provider?
Not anymore. The old way meant separate subscriptions and integrations for each provider, which was expensive and complex enough that many teams gave up and used one source.
Platforms now aggregate the providers for you. Tools like Clay bring many data sources together in one place, so you pay for what you use and configure the cascade in one system instead of managing a pile of contracts. That shift is what made waterfall enrichment practical for smaller teams.
The tool still isn't the strategy. Expecting the platform to do the thinking is one of the reasons most Clay implementations fail, because the provider choice and logic still decide the result.
Running waterfall enrichment on a cadence
A common misconception is that enrichment is a one-time job. Data decay says otherwise: people change jobs, companies move, and details shift constantly, so even a perfectly enriched record degrades over time.
Job changes alone keep that clock running. Bureau of Labor Statistics figures put median employee tenure in the US at 4.1 years in January 2026, so a meaningful share of any contact list moves on every year.
Finding data isn't the same as confirming it's still valid, so enrichment and verification work together. Repeating the cascade on a cadence, rather than once, is what keeps records complete and current as they age.
Most waterfalls that underdeliver were set up with one of these mistakes, and we keep finding them in the builds we take over:
Poor ordering wastes the most money. Putting an expensive or low-hit provider first slows resolution, when a smarter order would resolve most records cheaply.
Too many providers comes next. Beyond a handful, extra sources add cost without meaningfully improving coverage, so more sources isn't always better data.
Skipping verification undermines what you collected. A waterfall that gathers without verifying can fill records with stale or wrong information.
Treating it as one-time lets the gains rot. Without repeat runs, decay erodes your coverage until a list that was complete becomes unusable.
Recap
No B2B data provider has it all, and that isn't going to change. Chaining sources is the only way past any one provider's blind spots, and done well it raises coverage while keeping what you pay per filled field down.
The concept fits in a sentence. The results come from the configuration: which providers you pick for each field, the order they run in, when the cascade stops, and whether what it finds gets verified and refreshed.
The payoff runs through everything that consumes data. A waterfall is what makes data enrichment with firmographic data or technographic data close to complete, it turns list building into a list you can actually reach, and it lets lead scoring and ideal customer profile filtering work on full records instead of guessing from gaps. It's a foundational layer of any go-to-market system, especially once you automate sales prospecting at scale.
If you'd rather not find the right provider order by trial and error, we can set the cascade up in the Clay account you already have, test it against a sample of your list, and leave you with tables your team can read and change.
Frequently asked questions
What is the difference between waterfall enrichment and data enrichment?
Data enrichment is the general practice of adding missing details, such as emails, phone numbers, job titles or company size, to records you already hold. Waterfall enrichment is one way of doing it. Instead of pulling those details from a single provider, it queries several providers in a set order and only moves to the next one when the previous source finds nothing. The goal is the same, with more sources and higher coverage.
What data do you need to start waterfall enrichment?
For a work email, the usual starting point is the person's full name plus their company domain, and a LinkedIn profile URL makes matches more reliable. Clay's email waterfall, for example, treats the company domain as the primary input, recommends adding personal LinkedIn URLs, and offers full name plus company name as a fallback. For company data, a clean domain is usually enough. The cleaner your inputs, the more each provider in the chain can find.
Which tools offer waterfall enrichment?
Data orchestration platforms such as Clay are the best-known option. They connect many providers in one workspace, let you set the order for each field, and bill through credits instead of separate contracts. Sales databases such as Apollo also offer a waterfall that checks connected third-party sources, in an order you set, until one returns an email or phone number. You can also build one yourself with provider APIs and an automation tool, at the cost of more maintenance.
How is waterfall enrichment priced?
There is no standard price. Most platforms bill in credits, and harder fields cost more: FullEnrich, for example, charges 1 credit for a work email and 10 for a mobile number. The bigger question is whether empty lookups cost money. Clay refunds credits when a provider finds no email, while Apollo's documentation says some connected sources charge only when they return data and others may charge even when they find nothing. Check this per provider before you set the order.
What is the difference between waterfall and parallel enrichment?
Parallel enrichment sends every record to all providers at once, then picks or merges the answers. Waterfall enrichment asks providers one at a time and stops when one returns good data. Parallel can be faster and lets you compare sources against each other, but you query every provider for every record, which usually costs more. A waterfall is typically cheaper because later, pricier providers only see the records earlier ones missed.
Is waterfall enrichment GDPR compliant?
The method itself is neutral, but the data it collects is personal data, so the usual rules apply. You need a lawful basis, often legitimate interests for B2B outreach, and you must tell people you hold their data. The UK regulator's guidance on using data brokers says simply accepting a broker's assurance that its data is compliant is not enough, so check every provider in your chain.
What happens to records the waterfall cannot enrich?
They stay empty, and they should be marked as not found rather than left blank, so nobody mistakes them for records that were never processed. Then decide what is worth another try. Retrying with a different input, such as a LinkedIn URL instead of a name and company, can produce a match the first pass missed. High-value accounts can go to manual research. The rest can wait for the next scheduled re-run, since providers keep updating their data.
How do catch-all emails affect waterfall enrichment?
A catch-all domain accepts mail for any address, so a validator cannot confirm that a specific mailbox exists. That matters in a waterfall because the cascade stops once an email counts as valid. In Clay's email waterfall, catch-all emails are treated as valid by default. If deliverability matters more than volume, change that setting so catch-alls count as invalid and the waterfall keeps looking for a safer address.
Can waterfall enrichment run automatically on new inbound leads?
Yes. You can trigger the cascade from a form submission, a new CRM record or a webhook, so it runs without anyone pressing a button. Two things change compared with batch work. Each provider is tried in turn, so results take longer than a single lookup, and routing should wait until the waterfall finishes. You also need entry rules, such as skipping free email domains or existing customers, so automatic runs don't burn credits on junk.
Should waterfall results overwrite data already in your CRM?
Not by default. Set enrichment to fill empty fields only, then allow overwrites field by field where fresher data clearly wins, such as a job title after a job change. Protect fields that reps edit by hand and anything tied to consent. Check every route data takes into the CRM, too. In HubSpot, for example, the Enrich record workflow action can overwrite properties regardless of the overwrite rules you set for each property.
Share this post
Enrichment
Data
Glossary
Related glossary items
Understanding the core GTM concepts
Ready to build your pipeline
