Data Enrichment
What is waterfall enrichment? Why one data provider isn't enough
Waterfall enrichment explained: how it works, why it beats a single provider, how to order providers, control costs, verify results and build one in Clay.
Founder, OretonGTM · ·6 min read
If you’ve ever run a list through a single data provider and found emails for only part of it, you’ve run into the core limitation of B2B data: no provider has everyone. Each one is built from different sources, with different strengths by region, industry and company size.
Waterfall enrichment is the answer. Instead of relying on one provider, you query several in sequence until one returns a valid result. This guide explains how it works, why it matters, how to design one and how to keep costs under control.
How waterfall enrichment works
For each contact (say, name + company domain):
- Ask Provider A for the email.
- If A returns a result, verify it. If it’s valid, stop.
- If not, ask Provider B.
- Continue until you get a valid email or run out of providers.
Contact → Provider A → found? → verify → valid? → done
↓ no ↓ no
Provider B → found? → verify → valid? → done
↓ no
Provider C → ...
↓ no
Mark "not found"
The same logic works for phone numbers, company data (headcount, revenue, tech stack) and other fields.
Why one provider isn’t enough
Every data provider builds its database differently: some crawl the web, some rely on contributor networks, some aggregate public records, some infer email patterns. That leads to very different coverage:
- Geography: a provider strong in the US may be weak in Brazil or Germany.
- Company size: some providers cover enterprises well and startups poorly, or vice versa.
- Industry: technical and tech-adjacent industries are often better covered than traditional ones.
- Freshness: some update frequently; others lag behind job changes.
The result: the provider that’s best for one segment may be mediocre for another. A waterfall lets you combine their strengths.
My take: The question isn’t “which data provider is best?” It’s “which combination of providers gives me the best coverage for my ICP at an acceptable cost?”
What you can enrich with a waterfall
| Data point | Notes |
|---|---|
| Work email | The most common use; always verify |
| Mobile / direct dial | Expensive; enrich only for accounts you’ll call (find direct dials) |
| LinkedIn URL | Useful for matching and multichannel |
| Job title / seniority | To confirm the person is still in role |
| Company firmographics | Headcount, revenue, industry (firmographic data) |
| Technographics | Tools in use (technographic targeting) |
| Funding data | Stage, last round (funding signals) |
Designing a waterfall: step by step
Step 1: Define what you need
For most outbound: a verified work email for each contact, and a mobile number only for Tier A accounts.
Step 2: Test providers on your ICP
Don’t rely on vendor coverage claims. Take a sample of contacts from your actual ICP (a few hundred is a reasonable start), including a mix of regions and sizes you target, and run them through each candidate provider separately.
Measure for each provider:
- Find rate: share of contacts with a returned email
- Valid rate: share of returned emails that pass verification
- Cost per valid email
- Overlap: how many of its results others also found
Step 3: Order the providers
Common ordering strategies:
- Cost-first: cheapest providers first, expensive ones as fallbacks. Minimizes cost.
- Accuracy-first: most accurate providers first. Minimizes bad data, may cost more.
- Segment-specific: different orders for different regions or segments.
A typical approach is to put a strong, affordable provider first, then providers that cover the gaps the first one leaves.
Step 4: Verify every result
Providers return emails that can be outdated or pattern-guessed. Run each result through a verification service before accepting it. Results usually fall into:
- Valid: safe to send
- Invalid: discard and try the next provider
- Catch-all / risky: the domain accepts all mail, so the mailbox can’t be confirmed
Handle catch-alls separately: send in smaller batches, use additional validation, or deprioritize. See email verification and catch-all emails.
Step 5: Store the source
Record which provider found each email and when. This lets you measure provider performance over time, debug bounces and re-verify old data.
Building a waterfall in Clay
Clay is the most common tool for waterfalls because it connects to many providers and handles the “try the next one” logic for you.
A basic setup:
- Import your contacts or build them from a company list (see Clay tutorial).
- Add a waterfall enrichment for work email, choosing and ordering providers.
- Add verification as part of the waterfall or as a separate step.
- Add conditions so expensive steps only run for qualifying rows (e.g., only find phones if Tier = A).
- Output the final email, its source and its verification status.
- Send verified contacts to your sequencer and CRM.
You can also bring your own API keys for some providers, which changes the cost structure. See Clay credits and cost control.
Controlling costs
Waterfalls can be cost-efficient or expensive depending on how you run them.
Do:
- Qualify accounts before enriching people. Don’t find emails for contacts at companies you’ll disqualify.
- Enrich only what you’ll use soon. Data decays; enriching your whole TAM upfront wastes money.
- Stop at the first valid result. That’s the point of a waterfall.
- Dedupe first against your CRM and previous enrichment runs.
- Cache results so you don’t pay twice for the same contact.
- Run phones only for Tier A or accounts you’ll actually call.
Don’t:
- Put the most expensive provider first by default.
- Run every provider for every contact “just to compare”.
- Re-enrich contacts that were enriched recently.
Measuring waterfall performance
Track these per campaign and per provider:
| Metric | What it tells you |
|---|---|
| Overall find rate | Coverage of your waterfall for your ICP |
| Find rate by provider (incremental) | How much each provider adds after the ones before it |
| Valid rate after verification | Data quality |
| Bounce rate after sending | Real-world accuracy |
| Cost per verified email | Efficiency |
If a provider adds very few incremental results, consider removing it or moving it later. If bounce rates rise, check which provider sourced the bouncing emails.
Waterfall vs. single provider: when is one enough?
A single provider can be enough when:
- Your ICP is concentrated in a region and segment where one provider is clearly strong.
- Your volume is small and the extra coverage isn’t worth the complexity.
- You’re just testing outbound and need a quick start.
A waterfall is worth it when:
- You target multiple regions or less-covered markets.
- Coverage gaps are costing you meaningful parts of your TAM.
- Bounce rates from one provider are too high.
- You’re running outbound at scale and every percentage point of coverage matters.
See Clay vs Apollo and comparing data providers.
Compliance
Enrichment involves personal data. Make sure you:
- Use providers that are transparent about their sources and compliance.
- Collect only the data you need for outreach.
- Honor opt-outs and deletion requests across all systems.
- Understand the rules in your target markets (GDPR in the EU, LGPD in Brazil, CAN-SPAM in the US).
See is B2B data enrichment GDPR compliant? and is cold email legal?.
Common mistakes
- Trusting provider coverage claims instead of testing on your ICP.
- Skipping verification. Waterfalls find more emails, including more bad ones.
- Enriching everyone. Wasting credits on contacts you’ll never email.
- Not tracking sources. No way to know which provider causes bounces.
- Static order forever. Provider quality changes; re-test periodically.
- Treating catch-alls as valid. They can inflate bounce rates.
Where to start
Pick two or three providers, test them on a few hundred contacts from your ICP, order them by cost and incremental coverage, add verification, and run your next campaign through the waterfall. Compare find rate and bounce rate with your previous single-provider campaigns. That comparison usually makes the case on its own.
Frequently asked questions
What is waterfall enrichment?
A method of enriching contact or company data by querying multiple data providers one after another. If the first provider doesn't return a valid result, the next one is tried, and so on, until a result is found or the list of providers ends.
Why use waterfall enrichment instead of one provider?
No single provider has complete coverage. Each is stronger in certain regions, industries or company sizes. A waterfall combines their coverage, finding more valid contacts while only paying for the lookups you need.
How do I order providers in a waterfall?
Test providers on a sample of your actual ICP, measure find rate and accuracy, then order them by a mix of cost and performance: usually cheaper or best-performing providers first, more expensive ones as fallbacks.
Do I still need email verification with waterfall enrichment?
Yes. Providers can return outdated or guessed emails. Verifying every email before sending protects your bounce rate and sender reputation.
Is waterfall enrichment expensive?
It can be cost-efficient, because you stop at the first valid result and only pay for successful lookups with many providers. Costs grow when you enrich contacts you'll never use or include expensive providers early in the sequence.
Which tools support waterfall enrichment?
Clay is the most common tool for building waterfalls across many providers. Some data platforms also offer built-in waterfalls, and teams can build custom ones with n8n or code calling provider APIs.
Can I use waterfall enrichment for phone numbers?
Yes, and it's especially useful for mobile numbers, where coverage is patchy. Phone lookups are usually more expensive, so only run them for accounts you'll actually call.