What Is Waterfall Enrichment?

If you are active within the context of B2B data, you have heard the term waterfall enrichment. But what is it, and how has it become the industry standard?

Working with B2B data is challenging. Lead databases are incomplete, serve wrong data, or trap users with unexpected price hikes. For years now, it has become increasingly easy to become a B2B data provider.

Scraping providers like Bright Data, Coresignal, or Crustdata offer their data as resellable data bricks. New providers purchase one or more of these data bricks, add their own brand and marketing on top, and call themselves a data provider.

Consumers are left to derive strategies to turn unreliable systems into the fuel that powers their GTM operations.

One of these strategies is to chain multiple providers together in the hope that if the first provider won't find the data, the second will.

Waterfall

Waterfalls make use of the way B2B data providers bill. Most providers will only charge for an email address or phone number if it is found. No result. No charge.

This is an open invitation to chain many providers together without meaningfully increasing the cost. All we drive up is the round-trip time. The more providers we ask, the longer we may need to wait.

Waterfalls come in different flavors, and different providers sell them at different price points.

This blog post aims to serve as your map to navigate the market, so you can make an informed buying decision.

Waterfall schemas

The perfect waterfall

Only a few providers scrape the data themselves. It is much more common to purchase enrichment data from a vendor that purchased it from another vendor.

But that raises the question of redundancy.

If you chain vendors A and B, but both resell dataset C, your waterfall is pointless. Both providers ask provider C.

A perfect waterfall chains n providers together. In this chain, every provider contributes a meaningful amount of exclusive data to your waterfall.

Building such a waterfall requires careful benchmarking. Something that almost no provider in the wider enrichment industry does.

The foul waterfall

If a perfect waterfall is a chain of n providers where each provider adds meaningful exclusive data to the waterfall, a foul waterfall is a waterfall that chains an absurd number of providers with overlapping datasets together. This waterfall will waste a lot of your time, but you'll hardly gain anything by chaining these providers.

Waterfall offerings

The embedded waterfall

Over the last few years, several providers have specialized in providing waterfalls. A few notable mentions are Surfe, FullEnrich, and BetterContact.

The providers will manage a waterfall for you and charge a flat fee for each resolved value. Typically, this is anywhere from 30¢ to 100¢ per phone number and 8¢ to 30¢ per work email.

Typically, these providers all follow the same playbook. They resolve contact information from a LinkedIn URL. Before calling the first provider, they resolve the LinkedIn profile of the user, retrieve their profession and location, and, based on this data, adjust the final waterfall.

The frontloaded embedded waterfall

This strategy is a cost optimization over the embedded waterfall. Here, we're placing a fast and cheap provider before an embedded waterfall. This provider is often part of the embedded waterfall itself. By placing it before an embedded waterfall, we don't gain accuracy or coverage. Instead, we save roughly 75% of the cost for the 60% of values that are easy to resolve. Only for our hardest-to-resolve values do we call the most expensive embedded waterfall.

How to evaluate waterfalls

We've previously written an entire blog post about how to evaluate data enrichment providers.

The core challenge when doing so is to create and run benchmarks that tell you whether:

• A provider contributes meaningful unique data to your waterfalls
• The data that a provider returns is accurate

What waterfalls cannot do

A common misconception is that waterfalls increase accuracy. This is not true, or at least it is a more complex assessment.

Generally, waterfalls are a way to increase coverage, not accuracy.

Share notes with us

If you're looking for a new waterfall provider, contact us to get access to our benchmarking data.