Waterfall Enrichment: Maximize Data Coverage with Multiple Sources

What is Waterfall Enrichment?

Here’s a scenario you’ve probably experienced: You upload 1,000 leads to ZoomInfo to find their email addresses. It returns matches for about 650 of them. That’s a 65% match rate, which sounds decent until you realize you’re missing contact information for 350 potential customers.

What happens to those 350? In a traditional single-source approach, they’d just sit there as dead leads in your CRM. But waterfall enrichment changes the game entirely.

Waterfall enrichment is exactly what it sounds like: when one data source doesn’t have the information you need, you automatically query the next source in your sequence, then the next, and the next, until you either find a match or exhaust your options. It’s like having a backup plan for your backup plan.

Think about how a single-source approach works. You query ZoomInfo for an email address. If it finds the contact, great—you use that data. If it doesn’t find them, you’re done. That’s it. You’ve hit a wall, and roughly 30-40% of your leads are now unusable for outreach.

Now consider the waterfall approach. You start with ZoomInfo, and if it finds the contact, perfect—you stop there and use that data. But if ZoomInfo comes up empty, the system automatically queries Apollo. Found it there? Excellent, stop and use Apollo’s data. Still nothing? Move on to Hunter. Continue this process through your sequence of sources until you either find what you need or run out of options.

The beauty of this approach is simple: no single database contains everything. Each provider collects data differently, focuses on different company sizes, excels in different geographic regions, and has varying levels of database freshness. By combining multiple sources, you’re playing to each provider’s strengths while compensating for their weaknesses.

Why Waterfall Works: The Coverage Problem

Let’s talk numbers. When you rely on a single enrichment provider, you’re typically looking at coverage rates somewhere between 55% and 70%, depending on the provider and your target market. That means for every 100 leads you try to enrich, you’re losing 30 to 45 of them right off the bat.

Here’s how different providers typically perform in terms of coverage:

Now, why do these gaps exist? It comes down to a few key factors. Different providers use different data collection methods—some scrape websites, others rely on user contributions, and some use public records. They also have different geographic strengths. A provider focused on the US market might have terrible coverage for European contacts, and vice versa. Company size matters too. ZoomInfo tends to excel with enterprise companies, while Apollo has better penetration in the startup and SMB space. And finally, database freshness varies wildly. People change jobs, companies get acquired, and email patterns shift. No provider can keep up with all of it.

This is where waterfall enrichment becomes powerful. When you combine multiple sources, your coverage rates jump dramatically:

Let’s walk through a real example. You start with a list of 1,000 contacts and run them through ZoomInfo first. You get a 65% match rate—650 contacts found, 350 still missing. Those 350 missing contacts now go to Apollo, which finds 60% of them—that’s another 210 contacts. You’re now at 860 total, with 140 still missing. Feed those 140 to Hunter, which finds 55% of them—another 77 contacts. You’re now at 937 contacts enriched from your original 1,000. That’s a 93.7% match rate, compared to the 65% you started with.

The difference between 65% coverage and 94% coverage isn’t just academic. If you’re working with a list of 10,000 potential customers, that’s the difference between being able to contact 6,500 people versus 9,400 people. That’s 2,900 additional opportunities that would have otherwise been lost.

How to Implement Waterfall Enrichment

There are several ways to set up a waterfall enrichment system, ranging from no-code tools to custom development. Let’s walk through each approach and when it makes sense.

Using Clay for No-Code Waterfall

Clay has become the go-to tool for waterfall enrichment, and for good reason. It’s specifically built for this use case and makes the process incredibly straightforward.

Here’s how it works in practice: You upload your list or connect it from another source like your CRM or a CSV file. Then you add enrichment columns for the data points you need—email addresses, phone numbers, job titles, whatever. Next, you configure your waterfall sequence, telling Clay which sources to query and in what order. Click run, and Clay handles all the API calls, logic, and fallback sequencing automatically. When it’s done, you export the enriched data or push it directly to your CRM.

A typical Clay waterfall setup might look like this: You create a column for work email. Step one tries Clearbit. If that comes back empty, step two tries Apollo. Still empty? Step three tries Hunter. If Hunter doesn’t have it either, step four tries Snov.io. The result is that you get the email address from whichever source finds it first.

The pricing model is pay-as-you-go, typically ranging from $0.01 to $0.10 per enrichment depending on the source and data type. The beauty is that you only pay for what you use, and because the waterfall stops as soon as it finds a match, you’re not wasting credits on redundant lookups.

Building Custom Code Solutions

For companies with high volumes or specific requirements, building a custom waterfall system gives you complete control. This approach makes sense when you’re processing tens of thousands of records regularly or need complex conditional logic that goes beyond what no-code tools offer.

A custom waterfall enrichment function might start by querying Clearbit first because it’s the most accurate, though also the most expensive. If Clearbit returns an email with high confidence (say, above 80%), you use it and stop there. If not, you move to Apollo as your second choice. Found something? Great, use it and stop. Still nothing? Try Hunter next, but only if the confidence score is above 70. As a last resort, you might use pattern guessing based on the company’s email format, but you’d flag that email for verification before actually using it.

The benefits of custom code are clear: complete customization of your logic, precise cost control through conditional routing, ability to implement complex decision trees, and seamless integration with your existing systems. The downside is that you need engineering resources to build and maintain it.

Using iPaaS Tools Like Zapier or Make

If you’re already using automation tools like Zapier or Make, you can build waterfall logic into your existing workflows. This works well for simpler sequences and lower volumes.

A basic iPaaS waterfall might trigger on a new contact being added to your CRM. Step one does a Clearbit lookup. Then you add a filter: was an email found? If yes, update the CRM and end the workflow. If no, move to step two and try Apollo. Another filter checks if Apollo found an email. If yes, update CRM and end. If no, try Hunter. One final filter, and then you either update the CRM with the result or flag the contact for manual research.

The limitation with iPaaS tools is that more complex waterfalls require more zaps or scenarios, which can get messy and expensive. Managing credits across multiple providers is also trickier. For simple two or three-source waterfalls, though, this can be a good middle-ground solution.

Manual Waterfall for High-Value Leads

For enterprise deals or high-value accounts where each lead is worth significant revenue, manual waterfall enrichment might actually be the right approach. You check ZoomInfo first. Not there? Try Apollo. Still nothing? Use LinkedIn to verify the person’s current role and domain, then use Hunter to find likely email patterns. Still coming up empty? Make an educated guess based on the company’s email pattern and use an email verification tool to check it.

This process might take 5-10 minutes per lead, but if each closed deal is worth six or seven figures, that time investment makes perfect sense.

Optimizing Your Waterfall Source Order

The order in which you query data sources matters tremendously for both cost and data quality. Here are the key factors to consider when setting up your sequence.

First, accuracy should be your primary consideration. You want your most accurate source first because bad data is worse than no data. A wrong email address bounces and might damage your sender reputation. A wrong phone number wastes your sales team’s time. So even if a source is more expensive, if it’s significantly more accurate, it should often go first.

Second, consider coverage for your specific segment. If you’re selling to enterprise companies, ZoomInfo’s enterprise focus makes it a strong first choice. If you’re targeting startups and tech companies, Apollo’s strength in that market makes it a better lead source. Match your source order to your ideal customer profile.

Third, factor in cost—but as a secondary consideration after accuracy and segment fit. Your cheaper sources should generally come later in the waterfall, where they’re handling the stragglers that your premium sources couldn’t find. This way, you’re using expensive sources only when they’re most likely to deliver value.

For a general B2B audience, a good waterfall sequence might be: Clearbit first for accuracy, Apollo second for broad coverage, Hunter third for email pattern-based finding, and Snov.io last as an affordable backup option.

For enterprise-focused companies, you’d want ZoomInfo first because of their enterprise strength, Clearbit second for quality, and Apollo third as a supplemental source.

For startup and tech company targeting, flip it: Apollo first for their strong startup data, Clearbit second for quality verification, and Hunter third for pattern-based backup.

If you’re focused on European markets, you’ll want to start with Cognism because of their GDPR-compliant data collection, then Apollo for broader coverage, and Hunter as a backup.

Here’s a comparison of how different sources stack up across key dimensions:

Tracking Performance and Optimizing Over Time

Here’s where most companies drop the ball: they set up a waterfall sequence and then never look at the data to see what’s actually working. This is a massive missed opportunity.

You need to track specific metrics for each source in your waterfall. What’s the match rate—what percentage of records does this source find? What’s the accuracy—what percentage of the data from this source is actually valid when you verify it? What’s your cost per match from each source? And critically, what percentage of matches are unique to this source—meaning no other source in your waterfall had that data?

Here’s what this might look like in practice. You process 1,000 records. Clearbit goes first and finds 580 of them—a 58% match rate. Of those 580, you later verify that 94% are accurate. All 580 are unique to Clearbit because it went first. The remaining 420 records go to Apollo, which finds 252 of them—a 60% match rate on this subset. Of Apollo’s matches, 88% verify as accurate. Here’s the key insight: only 25% of what Apollo found was unique. The other 75% were records that Clearbit would have found too, had you queried them again. The remaining 168 records go to Hunter, which finds 92 of them—a 55% match rate. Hunter’s accuracy is 82%, and only 9% of Hunter’s matches are unique.

What do you do with this information? If Clearbit has low unique matches, meaning other sources are finding the same records, you might move Clearbit lower in your sequence to save money. You’d still get most of the same coverage but at a lower cost.

If Hunter has high unique matches—meaning it’s finding records nobody else has—you might consider moving it higher in your sequence, even though its accuracy is lower. You’d just add a verification step after Hunter to maintain data quality.

If a source consistently shows low accuracy, don’t drop it entirely. Instead, add an email verification step specifically after that source to catch invalid addresses before they go into your CRM.

Cost Optimization Strategies

Waterfall enrichment can get expensive if you’re not careful, but there are several ways to optimize costs while maintaining coverage.

The most important rule: stop on match. As soon as any source finds what you’re looking for, stop the waterfall. Don’t keep querying additional sources just to see what they might have. This seems obvious, but you’d be surprised how many custom implementations miss this step.

Use conditional routing based on what you know about the lead. If the company has more than 500 employees, start with ZoomInfo because that’s where they excel. For smaller companies, start with Apollo. This simple branching logic can significantly reduce wasted queries.

Batch similar records together to optimize routing. If you’re enriching a list of software engineers at tech startups, route them all through Apollo first. If you’re enriching C-suite executives at Fortune 500 companies, route them through ZoomInfo first. Don’t treat every record the same.

Cache your results and avoid re-querying records you’ve enriched recently. If you enriched a contact three weeks ago, you probably don’t need to enrich them again today. Set up a reasonable cache period—maybe 90 days for contact information, longer for relatively static company data.

Avoiding Common Waterfall Mistakes

Through working with hundreds of companies, we’ve seen the same mistakes repeated over and over. Here’s how to avoid them.

Don’t use too many sources. Companies get excited about waterfall enrichment and add 10 or 12 sources to their sequence. The problem is that after about four sources, you hit diminishing returns hard. Each additional source finds fewer unique records while still costing you money and time. Stick to three or four well-chosen sources.

Get your source order right from the start. Using cheap, inaccurate sources first might seem like it saves money, but it backfires. You end up with bad data in your CRM, your emails bounce, your sales team wastes time on wrong numbers, and you damage your sending reputation. Order by accuracy first, then cost.

Always verify, especially for email addresses. Even if you trust a source, verification is cheap (fractions of a cent per email) and protects you from bounces. Don’t skip this step.

Track your metrics. Without data on what’s working, you’re flying blind. Set up tracking from day one so you can optimize based on real performance rather than assumptions.

Finally, don’t over-engineer your waterfall. Start simple—maybe just two sources—and add complexity only if you need it. A simple waterfall that actually gets implemented is infinitely better than a complex one that stays in the planning phase for months.

Waterfall for Different Data Types

While we’ve focused mostly on email enrichment because that’s the most common use case, waterfall works for other data types too.

For phone number enrichment, a typical sequence might be Cognism first (they specialize in verified phone numbers), ZoomInfo second if you have access to it, Lusha third as an affordable option, and Apollo last as a supplemental source.

For company data enrichment—things like employee count, revenue, industry, and tech stack—you might use Clearbit Company API first, Apollo company data second, Crunchbase third specifically for funding information, and LinkedIn’s company page as a final fallback.

The principles remain the same regardless of data type: order by accuracy and segment fit, stop on match, and track performance to optimize over time.

Key Takeaways

Waterfall enrichment is one of the most effective ways to maximize your data coverage without breaking the bank. By querying multiple sources in sequence, you can boost coverage from the typical 60-70% you’d get from a single provider to 90% or higher.

The key is to be strategic about your approach. Order your sources by accuracy first, then coverage for your segment, then cost. Stop querying as soon as you find a match to avoid wasting credits. Use purpose-built tools like Clay for the easiest implementation, or build custom code if you need more control.

Track your performance metrics religiously. Know what each source is contributing in terms of match rates, accuracy, cost, and unique matches. Use this data to continuously optimize your waterfall sequence.

And remember: start simple. A two or three-source waterfall that you actually implement today will serve you better than a complex five-source system that you’re still planning next month.

The goal isn’t perfection—it’s maximum coverage at optimal cost, so you can spend less time hunting for contact information and more time actually connecting with potential customers.

Need Help With Waterfall Enrichment?

Setting up an effective waterfall enrichment system requires expertise across multiple platforms, understanding of data quality nuances, and ongoing optimization. We’ve implemented waterfall systems for hundreds of companies across various industries and use cases.

If you want to maximize your data coverage while controlling costs, book a call with our team. We’ll analyze your current approach, recommend an optimal waterfall sequence for your specific needs, and help you implement a system that actually works.