Resonate Append - Fill Rates, Match Rates, and Waterfall Matching

This guide explains how Resonate Append connects your records to our identity graph, why match rate and fill rate matter, how individual vs. household precision affects outcomes, and how our waterfall matching method balances quality and coverage.

What you’ll need and how to optimize:

• What’s needed: A Customer ID plus at least one other accepted identifier.
• Accepted IDs: HEM, MAID, ZIP11, Postal Address, IP Address.
• You can tune for precision or coverage (match rate) by varying the types and quantity of IDs you send, and we provide guidance on this below.

Understanding individual v. household matches

Different identifiers resolve to different levels of precision—and that affects targeting, segmentation, and measurement. Understanding whether your appended dataset needs to be at the individual or household level is critical, since it defines the granularity of the insights you’ll be able to apply.

• Individual-level IDs pinpoint one person and include HEM (hashed email), MAID (mobile ad IDs). Individual-level predictions offer the highest precision because they’re built from a single person’s behaviors, preferences, and intent signals, which enable more effective targeting, smarter segmentation, more relevant personalization, and greater confidence in predicting actions.

• Household-level IDs point to a shared home or device used by people at the same address and include ZIP11, Postal Address and IP Address. Household-level predictions are built on shared or inferred behavior across people living at the same address. While still useful they may lead to a more generalized messaging and slightly lower accuracy.

We maximize your addressable market by matching at both the individual and household level. That means you reach more of your target audience with precision.

Some matches are household level which means you reach the right home even if we cannot pinpoint which person in that home. For most marketing use cases that is exactly what you want because households make joint purchase decisions. The key is that we tell you which is “direct” and which is “household,” so you always know what you are getting.

But maximizing your reach and precision isn’t the only benefit gained with Resonate, you also get the most complete coverage. Let’s take a look.

Understanding fill rate and match rate

Two terms underpin the rest of this article:

• Match Rate: The number of records in your file that matches Resonate or other provider’s network of IDs.
• Fill Rate: The percentage of data that can be fulfilled by Resonate or other providers against your file, such as a CRM file. Fill rate may also be known as data coverage or data density. Resonate provides a 100% fill rate on all attributes for all matched IDs.

Think of match rate as "how many IDs we can connect,” and fill rate as “how complete those connected records are.” Together, these determine the value of your appended dataset.

Why this matters

Because we prioritize person-level first and clearly label if we resolve at individual or household level, what comes back isn’t just matched—it’s usable.

In this space, a high match rate can hide a low fill rate. The question to ask is: Of the records you matched, how many return every attribute I requested? That’s fill rate.

Here's an example: You send 50,000 records. 

• Resonate: we match 40,000 and deliver 100% fill—40,000 are usable. Usable coverage = 80% of your file. No blanks. No gaps.
• Others: they claim 45,000 matches, but if only 50% are fully populated, that’s 22,500 usable—45% usable coverage.

So yes, they ‘matched’ more—but you can use far fewer. We optimize for actionability, not vanity metrics. Every matched record—Individual or Household—is complete and ready to target, measure, and optimize.”

How our precision-focused matching works: Waterfall matching method
 

What is Waterfall Matching?

We use a waterfall method, starting with the most precise ID, then cascade to the next-most precise ID, prioritizing individual-level IDs before we consider household-level IDs. This process ensures we’re matching as many records possible, while still delivering the highest match quality.

We start with a direct match pairing your provided ID to the corresponding ID in our database. If there isn’t an available match at that level, we move down the line to other valid individual-level identifiers. And from there, we proceed to more approximate, household-level matching. 

Here’s a simplified example assuming HEM as the identifier:

By starting with the most accurate individual-level identifiers and moving to the next-best, we ensure you get the strongest possible connection—not just a convenient one.

We also believe you should have full transparency and visibility around our matching, which is why our output indicates the household or individual resolution for matching. This also helps validate accuracy and explain any insight expansion that may occur.

Why our waterfall matching process approach matters

Though some providers take a “match at all costs” approach, Resonate focuses on what matters most: quality, reliability, flexibility, and actionability.

Our waterfall matching method ensures we use the best available ID in your file, always prioritizing accuracy without discarding valuable matches. No matter what data you have, Resonate ensures you get the strongest possible match—and insights you can trust.

Input guidance for match rate and precision

Resonate’s waterfall matching requires a Customer ID + at least one of the other accepted ID types (HEM, MAID, Postal Address, ZIP11, IP Address). Expected match rates can vary depending on the data IDs you provide. The type and quantity of identifiers you submit can help you balance match rate (coverage) and precision (exactness). Please note that the more IDs you provide, the higher your match rates, however, some attributes could resolve at the household-level. 

Use the following guidance for tuning outcomes to anticipate your match rate and precision based on your identifiers:

• HEM and/or MAID: For a direct match (40-60%) with most-precise, individual level predictions
• HEM + ZIP11 or Postal Address: Stronger match rate (60-80%) when a direct HEM match is not available with a mix of individual- and household-level predictions.
• Customer ID + as many identifiers possible: Highest overall match rate (60-80% or higher) and most complete coverage across individual- and household-level predictions

Even if you don’t have individual-level identifiers like hashed emails or MAIDs—Resonate still delivers strong results using household-level matching through addresses, ZIP11 or IP.

Want to improve your match rates?

To optimize match performance:

• Follow our Data Preparation Best Practices
• Submit as many qualifying identifiers as possible (HEMs, MAIDs, Postal Address, ZIP11, IP Address)
• Format your files correctly using our File Formatting Guidelines
• For HEMs, use our hashing script to standardize inputs before submission

Please note that  while additional IDs can improve your match rates, some attributes could resolve at the household-level.

Resonate continuously improves our data coverage and focuses on perpetually enhancing our identity graph, and we’ve seen many clients' match rates grow over time.   

Still have questions?

We’re happy to walk you through your match results and unlock the full value of your first-party data. Contact your customer success manager or reach out to resonatesupport@resonate.com.