Aggregated Event Measurement is Meta's post-iOS 14 attribution framework that allows up to 8 prioritized conversion events per domain, reported as aggregated counts after a 24–48 hour delay to comply with Apple's ATT requirements.
Aggregated Event Measurement (AEM) is Meta's privacy-compliant attribution system for iOS 14.5+ users who opted out of tracking under Apple's ATT prompt. Instead of user-level signals, it delivers domain-level aggregated event counts — with a 24–48 hour reporting delay built in to prevent fingerprinting.
The mechanism works at the domain level: in Events Manager, you verify your domain and assign a ranked list of up to 8 conversion events in priority order. When an iOS user who declined ATT triggers a conversion, Meta reports that event against the prioritized list — but only the highest-ranked eligible event for each user gets counted. Ad sets can optimize for exactly one event on iOS traffic, and only if that event sits within your top 8. This forces a hard architectural choice: if you run deep-funnel sequences (lead → trial → purchase → upgrade → referral), at least one event tier gets cut from optimization eligibility entirely.
For your day-to-day measurement, AEM is the reason you see iOS attribution trailing web by a full day in your dashboard. Real-time pixel data for non-iOS sessions runs alongside AEM's delayed aggregated counts — and without explicit lag adjustments in your reporting stack, the mismatch corrupts daily budget decisions. I've seen accounts pulling iOS data at 9am and making bid changes before the prior day's conversions have fully surfaced. That's an AEM timing problem, not a campaign problem.
In the current 2025–2026 paid-media environment, AEM's constraints interact directly with Advantage+ campaign structures. Meta's Andromeda delivery system consolidates at the account level, but AEM still operates per-domain — meaning conversion API (CAPI) server-side signals are your best lever for filling the iOS signal gap. AEM pairs with app tracking transparency (ATT) consent rates: higher ATT opt-in in your audience reduces the proportion of conversions running through AEM's delayed aggregation. And for teams investing in Meridian or Robyn MMM models, AEM-attributed iOS conversions require explicit 48-hour lag corrections before ingestion — otherwise your MMM weights iOS spend against under-counted outcomes.
For a broader view of how attribution has fractured across the ecosystem, the post on death of attribution and marketing measurement in 2026 covers the landscape. If you're evaluating measurement tooling, ai analytics tools for marketing in 2026 benchmarks several platforms on their AEM handling.
Treat your AEM priority list as a live config, not a one-time setup.
AEM is why your iOS reporting trails web by a day, why you can only optimize for one event per ad set on iOS traffic, and why event prioritization decisions made in 2021 are still silently shaping your account performance right now. Most accounts I've audited have stale AEM configurations — event #1 still reflects an old funnel, the optimization event the ad set uses is actually #3 on the list, and nobody has reviewed it since iOS 14 launched. That silent misconfiguration costs you attribution accuracy and conversion volume at the same time.
