Docs

iOS 27.2's EU ATT Prompt: Tracking Annual Re-Consent

Apple's iOS 27.2 alternative ATT prompt and annual re-prompting in five EU countries change opt-in rates over time — the events to track before they do.

Acquisition

Apple’s iOS 27.2 update, rolling out to Germany, France, Italy, Poland and Romania following antitrust pressure from EU regulators, changes App Tracking Transparency in two ways that matter for attribution: an alternative system prompt with softer formatting and an optional explanatory link, and the ability to re-prompt a user for consent once a year regardless of their original choice. Both changes were announced in the run-up to this release and are now shipping in the affected markets.

The immediate risk is treating opt-in rate as a single, stable number. It won’t be. The alternative prompt’s wording and layout differ enough from the original binary “Allow” / “Ask App Not to Track” screen that opt-in rates in the five EU markets will shift independent of anything the app did — a change in Apple’s UI, not in the app’s targeting or trust. Layer annual re-prompting on top and a chunk of the install base will pass through consent twice in a year, at different points in their lifecycle, potentially reversing an earlier decision. Attribution models built on “IDFA present or not” as a fixed per-user flag will start seeing users flip state mid-cohort, and any LTV or ROAS model that assumes consent is captured once at install and never changes will quietly misattribute revenue to the wrong cohort.

Data Points to Track

  • Prompt variant shown (standard vs. alternative EU prompt) alongside every consent event, since the two aren’t comparable in aggregate reporting
  • Consent event with a re-prompt flag, distinguishing a user’s first ATT decision from a subsequent annual re-prompt
  • Prior vs. new consent state on every re-prompt event, to catch reversals in either direction
  • Country/region at time of prompt, since only five EU markets get the alternative flow — everyone else still sees the original
  • Time-since-install at each consent event, to separate install-time opt-in behaviour from re-prompt behaviour a year later
  • Attribution window validity, flagging any attribution record whose consent state has since changed

Setup Steps

  1. Capture the prompt variant (standard or alternative) as a property on the ATT consent event, not just the resulting permission state.
  2. Add a re-prompt boolean to the consent event schema so annual re-consent doesn’t get counted as a fresh install-time opt-in.
  3. Log the previous consent state before overwriting it, so a reversal from allow-to-deny (or the reverse) is an explicit, queryable event.
  4. Segment opt-in rate reporting by country and prompt variant rather than blending EU and non-EU markets into one global rate.
  5. Flag attribution records tied to a user whose consent state has changed since the attributed event, so downstream ROAS calculations can exclude or re-weight them.

Actionable Insights

The number to watch isn’t the blended opt-in rate — it’s the delta between the alternative prompt’s opt-in rate and the standard prompt’s, and separately, the reversal rate on annual re-prompts. If the alternative prompt lifts opt-in meaningfully in the five EU markets, that’s a real gain in attributable install volume worth reflecting in EU campaign budgets. If annual re-prompting produces a high reversal rate from allow to deny, it signals that the original consent was fragile — users who said yes once but didn’t feel strongly about it — and models trained on year-one attribution data should be revalidated once re-prompt data starts flowing in.

Expert help

Need help tracking this in your app?

Our team sets up analytics pipelines for mobile and web teams every day. Talk to us and get your first events flowing in under an hour.

Talk to an expert