Demystifying Match Rates: Getting to what really matters

(The following is an excerpt from @LiveRamp's eBook exploring match rate)

The emergence of data collaboration and data clean rooms , the availability of new apps, and the rise of AI introduce more opportunities to build meaningful data to drive decision-making, personalization, and revenue.?

At the same time, however, you face more data complexity. Within your company, you have data silos making it difficult to connect a complete picture of a person. And to make matters worse, increased privacy regulations, signal loss from third-party cookie deprecation, and decreased availability of mobile ad identifiers (MAIDs) and identifiers for advertisers (IDFAs) hinder your ability to connect your customer data across the Ad-Tech ecosystem.

In short, new technologies create data silos, making it hard to get an accurate customer view.?

Ultimately, to realize your data’s full potential, you need to be able to easily resolve, activate, and measure it across any number of partners and platforms.

Enter match rates – the leading indicators of how effectively you’re connecting your data across your ecosystem and reaching your customers. But when it comes to match rates, it seems like everyone has a different definition. So let’s break them down.

Match rates – what are they really?

In digital advertising, “match rate” refers to the percentage of a marketer’s target audience matched with a specific advertising platform. Put another way, it’s the number of your customers that exist as a targetable audience on the platform.?

Outside advertising, “match rate” might refer to the percentage of a file you can successfully overlap with another vendor or partner universe (e.g., a data seller spine).?

“Match rate” can also refer to the ability to recognize onsite visitors or even the deduplication or consolidation of their first-party data .

All these definitions are correct, from a certain point of view. Yet, within most of these definitions, there’s a core dynamic being balanced continually – maximizing scale and reach means maximizing match rates. But maximizing accuracy can often mean being more exacting about what qualifies as a match, which can lead to a lower match rate.?

Here’s an easier way to think about it:

• If you want more personalized targeting, focus on accuracy.?
• If you want to reach as large of an audience as possible, focus on scale.?

So how do you navigate the different definitions of “match rate” to drive the right outcomes?

It all starts with Identifiers—the keys used to match on.?

Different identifiers bring differing levels of match accuracy, reach, and precision. Here are some common ones:

Hashed emails (HEMs): A method of encrypting an email address by giving it a unique 32-, 40-, or 64-character code. This code remains the same no matter if the email address is used on different platforms, browsers, or devices. Many marketers use HEMs due to their low cost. But they come with limitations because they offer only one potential touchpoint to match on. HEMs also pose privacy concerns due to potential reidentification risks. With 86% of customers having multiple email addresses, you’ll lose potential reach with HEMs. Moreover, you can’t seamlessly protect your customer data because this same identifier, or key, runs the risk of being decrypted due to its large volume of usage.

Mobile advertising IDs (MAIDs): Unique, random alphanumeric identifiers that iOS or Android assigns to each mobile device. Although MAIDs can act as a device identifier tied to an individual, their availability has steadily decreased through technological changes. In addition, mobile devices (e.g., tablets) are often shared between household members, making it hard to personalize based on MAIDs alone.

First-party cookies: Identifiers stored by the website (or domain) you visit. These cookies allow website owners to collect analytics data, remember language settings, and perform other useful functions that help enable a good customer experience.

Third-party cookies: Identifiers placed on a customer’s device by a website from a domain other than the one the customer is visiting. Third-party cookies track a customer’s browsing history and activities, so they can present them with personalized ads for products and services. This identifier is only widely available in Google Chrome and is set to be deprecated soon, leading to potential signal loss. It’s just as well because relying on third-party cookies is risky business. In fact, Deloitte found that “companies across a range of industries risk an average of $91 million to $203 million in revenues per year due to the loss of third-party cookie data and signal loss and the resulting impacts on advertising effectiveness.

Digital identifiers: Unique identifiers for website visitors using a combination of first-party cookies and offline data related to those visitors. Digital identifiers don’t require marketers and ad publishers to sync cookies across multiple platforms. This limits inefficiencies and helps deliver the right ads to customers in the right context. However, as digital identifiers are often built on one or two pieces of known data, they typically don’t provide a true person-based view across multiple devices, emails, or phone numbers.

Enterprise identifier: A custom-encrypted identifier built from your identity rules and infrastructure. This allows you to control things like:

• Rules for when you add new records vs. merge records

• Access to new data sets to complete or append to records

• The ability to perform data hygiene

• The ability to resolve all touchpoints (e.g., multiple emails, addresses, phone numbers, devices) to a person-level ID – and identify a household for shared devices, differentiating between the individual customer and other members of the household

An enterprise identifier is unique to your business, giving you the ability to protect your customer identities because decryption cannot be completed internally or externally. It allows for seamless translation across multiple identifier types, enabling connectivity across the digital ecosystem to any number of platforms – while protecting customer privacy.

Each of the above identifiers can be useful, but here’s the key: To maximize your data connectivity and marketing effectiveness, you should:

• Maximize the number of identifiers you’re using.

• Work with a partner that can take in as many identifiers as possible and resolve all of them to a single, person-based view.

Matching levels—choose what you need

Now that we’ve looked at the different types of identifiers, the next thing to consider is: What level are you matching at? For example, are you matching at the individual level, household level, or another geo-level (e.g., ZIP+4)?

All have value – it just depends on your use case or outcome. Let’s take a closer look:

Identifier or device matching: Matches only when the exact identifier (e.g.,HEM, MAID, third-party cookie) matches another exact identifier. This is matching at its most basic, providing limited scale and accuracy due to only having one identifier to match on.

Person-level matching: Resolves multiple identifiers or known touchpoints to a single person based on persistent signals (whether deterministic or probabilistic) that refer to a single individual.

Household-level matching: Provides greater reach than person-level matching and matches multiple identifiers (or known touchpoints) to a single record of all the people determined to be in that person’s household. (At LiveRamp, we define customers as being in the same household if they reside at the same address and show a persistent relationship.) Household-level matching is especially useful for measurement use cases. For instance, one person in the household might see an ad and prompt another household member to purchase the advertised item.?

Neighborhood-level matching: Matches identifiers or known touchpoints to all the individuals determined to be in that person’s ZIP+4 area. Neighborhood-level matching is often used to maximize reach, especially for top-of-funnel tactics. The downside is less accuracy – and less personalized targeting.

With so many different ways to define “match rate” and the many factors involved, it’s no wonder confusion abounds. It’s easy to lose sight of the objective: What value are we trying to provide, and what business problems are we looking to solve?

So, what’s the answer? Get clear on the definitions and how they affect the outcomes you’re driving. It’s how best-in-class marketers reach their customers with the lowest friction and highest value.

The new @LiveRamp eBook, Demystifying Match Rates , contains even more important information and considerations, including a list of questions to ask when you're evaluating a match rate partner.

But the bottom line is this, wherever you are on the mach rate journey, LiveRamp offers solutions for every step. We can help you:

• Build your unique enterprise ID from multiple identifiers and achieve a true view of your customers across multiple systems, tools, and touchpoints.

• Leverage multiple match levels depending on your goals, according to your marketing tactics, to maximize return on your investment.

• Gain access to 400+ platforms and partners to seamlessly activate your data across the ecosystem.

• Translate to platform-specific IDs to maximize your ability to reach your intended audiences where they’re spending time.

• Collaborate with publishers and partners in data clean room environments to gain analytics insights with previously inaccessible data.

Check out the eBook and get in touch with us at LiveRamp if we can help!