Why Email is Not a Deterministic ID

By Jon Stewart January 29, 2021

Why Email is Not a Deterministic ID

Measurement of advertising outcomes relies upon deterministic IDs. If we are unable to isolate a person that was exposed to an ad from a person that was not, it is impossible to know whether the ad had any impact. Unfortunately, while valuable as a basis for targeting, email falls short as a basis for measurement. Why? Email is not a deterministic ID.
Years ago cookies were thought to offer deterministic identification at the person level, but over time the introduction of GDPR, the loss of browser cookie support, and growing mobile device usage eroded their value as a basis for person-level measurement. As cookies *ahem* crumbled, the industry rushed to find a new 1:1 identifier rather than focusing on a truly deterministic measurement solution.
The search for a new 1:1 identifier has led many in the industry towards hashed emails. As recently as last year, email addresses were considered by many to be the cookie replacement of the future. The reality, however, has proven to be quite different.

Issues with Email

The first issue is one of scale. It is estimated that only 10% of Internet traffic is logged in – leaving email identification blind to 90% of ad exposure online. Without accounting for 90% of traffic, emails offer no programmatic coverage and are unable to prevent control contamination. As the final nail in email’s proverbial coffin, Apple iOS14 terms clearly state that sharing emails of users who do not consent to IDFA can get you kicked off the app store.
However scale is not the only issue with email-based measurement. Due to widespread device sharing within households, it turns out the premise of person-level exposure data as the basis for deterministic measurement has been flawed from the start. It turns out that measurement has effectively been household-level all along. As a result, we have as an industry not only drawn the ire of privacy-conscious consumers and regulators, but we have done so needlessly, because person-level data only contaminates control groups thus depressing the observed effectiveness of advertising.
By recognizing the fact that 1:1 tracking has always been probabilistic, we can shift focus to a future-proof, privacy-safe solution for measuring ad effectiveness that delivers person-level exposure probability and deterministic, uncontaminated control groups. That solution is micro cohorts.

Advantages of Micro Cohorts

Micro cohorts are small groups of people, such as a household, a small geographical area or any collection of up to a few hundred people, that are measured as a privacy-safe surrogate for the underlying individuals. With adjustments for exposure probability, micro cohorts deliver the benefits of 1:1 measurement without the user-level contamination of test and control groups.
Feverishly constructing increasingly complex — and ultimately futile — person-level deterministic IDs for measurement is the way of the past. The future belongs to advertisers with the most meaningfully segmented audience cohorts and measurement models, not the largest collection of email addresses.
To learn more about micro cohorts, check out our recent Expert Update from Upwave, featuring insights from Upwave VP of Product Ken Archer, and MightyHive Sr. Director Myles Younger.