Key takeaways
- A waterfall asks provider A for a field, then provider B only if A comes back empty, and so on. You pay for coverage, not for duplicate answers.
- It earns its keep on emails, phone numbers and hard-to-find contacts. For basic firmographics, one good provider is usually enough.
- Enrichment only helps if the result is validated, mapped to the right CRM fields and written back without overwriting good data.
You exported 5,000 target accounts from the CRM, ran them through your enrichment tool, and a stubborn chunk of the contacts still have no email. So you buy a second tool, run the gaps through it, and fill some in. Then a third.
That manual process is a waterfall. Waterfall enrichment queries several data providers in a fixed order for the same field, stops at the first valid answer, and only pays for the providers it needed. Done well, it lifts coverage without paying twice for the same record.
How a waterfall actually works
Each provider has different strengths. One is good with US enterprise contacts, another with European SMBs, another with mobile numbers. Nobody covers everything.
A waterfall ranks them and runs them in sequence. If the first provider returns a verified email, the record is done. If not, it moves to the second, and so on.
Two details matter more than the provider list. The order, because it decides your cost. And the validation step, because a “found” email that bounces is worse than a blank.
Coverage versus cost
A waterfall isn’t automatically cheaper. It’s cheaper than buying every provider’s full plan and enriching everything with all of them. It’s more expensive than one provider if one provider would’ve been enough.
Here’s an illustration (not real pricing). Say you have 10,000 contacts missing an email:
- Provider A fills 5,000 of them.
- Provider B fills 2,000 of the remaining 5,000.
- Provider C fills another 800 of the remaining 3,000.
You end up with 7,800 emails instead of 5,000, and you only paid provider B for 5,000 lookups and provider C for 3,000. The last slice of coverage costs the most per record, so decide whether those contacts are worth chasing.
When you need a waterfall
Use one when the field is hard to find and the record is valuable.
| Situation | Waterfall? | Why |
|---|---|---|
| Work emails for outbound lists | Yes | Coverage varies a lot by region and seniority |
| Direct dials and mobile numbers | Yes | Single providers have thin coverage |
| Industry, headcount, location | Usually no | One reliable source is enough |
| Technographics (tools in use) | Sometimes | Depends on how specific the tech is |
| Inbound leads, form fills | Rarely | You already have the email, you need company data |
If your problem is that the CRM is full of duplicates and half-empty fields, enrichment won’t fix it. Clean first. Our CRM data hygiene checklist covers the order.
Not sure if a waterfall is worth it for your data?
Book a free 30-minute consultation. We'll review your CRM and target lists and show you how enrichment would work for your team, and where it wouldn't.
Setting one up without wasting credits
Most waterfall projects go wrong in the setup, not the providers.
List the three or four fields that block routing, scoring or outreach. Don't enrich everything.
Run 500 of your own records through each provider and measure fill rate and bounce rate.
Put the best accuracy per credit first, not the best brand name.
Verify emails and phone formats before anything touches the CRM.
Decide what can overwrite existing values. The default should be "fill blanks only".
Testing on your own records matters. Provider coverage claims are averages, and your territory and titles aren’t average.
What good looks like after the first month
Don’t judge the project by how many fields got filled. Judge it by what changed downstream. Are fewer outbound emails bouncing? Are reps spending less time hunting for contacts? Are accounts routing without manual fixes?
Here’s a simple illustration (not a client result). A 40-person sales team has 3,000 target accounts and each rep needs three verified contacts per account. Instead of asking reps to research, the waterfall runs overnight, flags the accounts where no provider found a senior contact, and drops only those into a manual queue. Reps get a smaller, higher-value research list, and the rest arrives in the CRM ready to use.
Where the data should land
Enrichment is only useful once it’s in the right place. A few rules we’d insist on:
- Match on domain first, not company name, so you don’t enrich or create duplicates. See our guide to Salesforce duplicate management for how matching should work.
- Store the source and the date on each enriched field. When something bounces, you’ll want to know which provider supplied it.
- Never overwrite a value a rep typed in by hand unless a rule says so.
- Re-enrich on a schedule. People change jobs, and a six-month-old email is a risk.
What enrichment feeds downstream
Better data is the input to everything else. A B2B lead scoring model can’t score company size if the field is empty. Routing rules can’t send an account to the right territory without a country. And if you’re planning AI agents on top of your CRM, our CRM AI readiness checklist is a sensible first read.
A reasonable rule: if a field drives a decision, it needs a source you trust. If it’s just nice to have, don’t pay to enrich it.
Mistakes we see most often
- Enriching the whole database once, then never again.
- Running three providers in parallel instead of in sequence, so you pay for all of them on every record.
- Skipping validation and sending to unverified emails, which hurts sender reputation.
- Letting enrichment overwrite clean data with worse data.
- Buying a bigger waterfall when the real problem is a messy CRM.
The bottom line
A waterfall is a coverage tool, not a data strategy. Use it for the hard-to-find fields, put validation in the middle, and write back carefully. If you’re unsure whether your data needs it, a free 30-minute consultation is a quick way to find out before you spend on credits.
Want this working for your team?
Book a free 30-minute consultation. We’ll review your current setup and use cases, show you how it would work for you, and leave you with a recommended plan.