How Aggregated Event Measurement works
When Apple began asking iPhone users whether apps could track them across other companies’ services, a large share said no. Meta could no longer report those people’s conversions the way it always had, so it built Aggregated Event Measurement — AEM — as the replacement. It reports totals without reporting individuals.
Three rules define it. First, each verified domain configures a limited set of web events, ranked in order of importance. Second, when someone triggers several of those events in a session, only the highest-ranked one is reported. Third, results arrive later than they used to, and are not broken down by age, gender, region or placement for the affected traffic.
The configuration lives in Events Manager, under the domain — not the ad account. That detail matters: if several ad accounts advertise the same website, they all share one event set, and whoever edits it changes measurement for everybody.
Why Aggregated Event Measurement matters
It sets the ceiling on what Meta can optimise towards. An event you did not include in the configured set is not available as a campaign objective for that traffic, however carefully the pixel fires it. Getting the set wrong quietly limits every campaign that runs against the domain.
It also explains reporting behaviour that otherwise looks like a fault. Numbers that climb for a day or two after the event, a breakdown that has gone blank, a conversion that shows in your own analytics but not in Meta — these are usually AEM working as designed rather than tracking that has broken.
Where it goes wrong
The most common failure is never configuring it at all. Domain verification is the prerequisite, and an unverified domain leaves the event set unclaimed, so anyone with access to the pixel can take it. Verification is a small technical job that is repeatedly left undone.
The second is a badly ordered set: a soft signal such as a page view or content view ranked above the purchase or lead you actually care about. Because only the top-ranked event survives, the outcome that pays your bills disappears from a share of your reporting.
The third is treating the configuration as permanent. Businesses add new forms, new checkouts and new offers, and the event set stays as it was set up years earlier.
What to do about it
Verify the domain first, in Meta Business Manager, and do it for the exact domain your conversions happen on. Then rank the set with your genuine money-making outcome at the top, and put softer signals lower down where they cost you nothing. Reserve room at the bottom for events you are still testing.
Expect delayed and partial reporting on the affected traffic, and judge campaigns over longer stretches rather than by yesterday’s figures. Feed events through the pixel and Conversions API together so browser-side loss does not compound the limits, and read platform results next to a source Meta does not control — your own sales records, or a properly configured GA4 setup.
