One year later: gauging the impact of Apple’s AppTrackingTransparency
Cross-app tracking is not dead, despite the far-reaching impacts of Apple’s AppTrackingTransparency framework. However, its use cases have changed, argues Jake Moskowitz, vice-president of data strategy at Emodo.
/ Adobe Stock
Many underestimate the impact of Apple’s AppTrackingTransparency (ATT) — which gives users control over which apps and sites can track their activity across other apps and sites on Apple devices. They note that it only impacts iOS and not Android users; they point out that it only applies to in-app, not web.
But the reality is that ATT was the first major move in the US of what is clearly a massive paradigm shift disrupting one-to-one-based targeting and measurement — with which, for many years now, digital marketing has been synonymous. Google’s plans with both Chrome and Android to remove opt-out tracking mechanisms spell the death knell for the way we’ve used digital for the last 15 years.
Much of the chatter in the industry has been focused on ways to reestablish one-to-one targeting and measurement via opt-in mechanisms such as first-party data and alternative identifiers. But opt-in can never replicate the scale of opt-out, as they are fundamentally different things. Even Bill Michels, executive vice-president of product and engineering at The Trade Desk, the creator of UID 2.0, said at an IAB Town Hall in November 2021: “On display or browser-based activity, we're not going to log into every website that we see... a guess there would be 20% of that is logged-in."
That’s not to say there’s not an important role for opt-in data such as first-party or alternative ID data, or that it’s not a worthwhile initiative for publishers and marketers to maximize the quantity and quality of their opt-in data. But it cannot serve as the primary means to make up for the scale loss as opt-out data ceases to exist. Its purpose is not just for one-to-one targeting and measurement, but rather also as training data for AI models. Only those AI models can make up for the scale loss. It’s time for marketers to shift their thinking entirely.
Has mobile measurement evolved since Apple introduced AppTrackingTransparency? The scary thing is, it’s hardly changed at all. Measurement providers that depend on one-to-one cross-app and cross-device tracking have barely adjusted — at least externally. Methodologies such as multi-touch attribution, sales lift, brand lift, in-store traffic lift, and more cannot exist in their current forms without mass scale user-level database matching of opt-out data. Marketers who depend on these methodologies should be actively exploring alternatives, or vendors that are taking the paradigm change seriously.
The one area in which we are seeing change is metrics that don’t require one-to-one user database matching, especially with the rise of attention metrics. They get a lot of press, and that makes sense given the promise that attention offers as a proxy metric for key lower-funnel metrics — a proxy that can exist in a world without massive scale of opt-out user tracking data. We still don’t have standards for what attention means or how to measure it, but expect growth to continue here. In fact, it’s not just attention that will pick up steam, but also other promising proxy metrics that don’t require user tracking.
While Google hasn’t expressly followed in Apple’s footsteps, the writing is on the wall. It has committed to removing third-party cookies from Chrome by the end of 2023. It has changed the opt-out nature of the Google advertising ID (ADID) on Android. Marketers should be actively evolving their targeting and measurement strategies.
Is cross-app tracking dead? Absolutely not. But the use case has changed. Opt-in methods can never replicate the scale of opt-out methods of tracking. And that means that opt-in methods can’t be relied on as the primary mechanism for one-to-one targeting and measurement in the way marketers have grown accustomed. But cross-app and cross-device tracking is important as ever as training data for AI models that can replicate the scale that marketers demand.
For brands who are future-proofing their mobile marketing strategies, the increasing focus on proxy measurement metrics such as attention that don’t require one-to-one user tracking is a good start. Many brands and publishers are also moving swiftly to maximize the scale of their first-party relationships with customers. This alone is not a future-proof strategy, but can prove especially effective in building up training data for AI models.
Ultimately, the time to act is now.
Jake Moskowitz is vice-president of data strategy at Emodo.