For an iOS app campaign on Meta, which attribution setup should we use (Aggregated Event Measurement or SKAdNetwork and AdAttributionKit), and how should we read the conversions Meta can't see?
Our answer
Use Meta's Aggregated Event Measurement (AEM) when your app and event are eligible, and send each event once. Keep SKAdNetwork as a separate view across networks. Judge both against App Store Connect and your revenue backend.
- Confidence
- Medium
- Strongest proof
- 2 Meta rules
- 1 dataset, 4 opinions or claims
- Sources
- 5 sources
- 7 statements
- MR
L
New here? Take the 1-minute tour
Why this is the answer
3 reasons and our judgmentApp promotion campaigns can now run on AEM without SKAdNetwork events, with near real-time reporting, 7-day click windows and Audience Network delivery (Meta). AEM doesn't yet support custom event optimization.
AEM and SKAdNetwork are two separate scoreboards (RevenueCat). AEM optimization needs no SKAdNetwork value update, the two results shouldn't be added together, and unmatched conversions never show as attributed.
Meta can miss a large share of iOS conversions even with AEM. Samet Durgun puts it at 30 to 60% and treats App Store Connect and your subscription backend as the source of truth (reported by Adapty).
Our judgment
- Make AEM the default for an eligible iOS app campaign that optimizes a supported event. Meta's page lists what AEM campaigns get without SKAdNetwork setup, and RevenueCat confirms no SKAdNetwork value update is needed. We take the leap from that to a default.
- When the event is custom or the app isn't eligible, fall back to the SKAdNetwork setting only after checking eligibility in Events Manager. Check eligibility in Events Manager first, as for the habit event in Optimization events for app campaigns. If it fails, map the event in SKAdNetwork, the other route RevenueCat describes.
- Keep SKAdNetwork or AdAttributionKit beside AEM when you buy iOS installs on more than one network. Use a single updater. Meta lets you keep configuring SKAdNetwork events for the view across platforms, and RevenueCat and Admiral Media treat Apple's frameworks as a separate path to keep clean.
- Judge the campaign on revenue in your own records, not on whichever system reports more. This applies Durgun's source-of-truth rule and RevenueCat's point that an event Meta accepts is not the same as one it attributes.
Our judgment, built from Meta, RevenueCat, Admiral Media and Samet Durgun: our reading of these sources, not a finding from any one of them.
What the sources say
7 statements from 5 sourcesMeta says Aggregated Event Measurement now covers app promotion campaigns, so they can run without configuring SKAdNetwork events. App events under it do not yet support custom event optimization.
“This means you can now run app promotion campaigns without configuring app events for Apple's SKAdNetwork”
Meta says you can rank up to 63 app events for SKAdNetwork, and only the highest-priority event is reported when several happen in a conversion window. Version 4.0 support is rolling out gradually.
“Only the highest-priority event is reported when there are multiple customer actions during a conversion window.”
RevenueCat support says Meta's AEM and Apple's SKAN are two separate scoreboards to compare, not add up, and that organic or unmatched trials will not show as attributed in Ads Manager. It gives no data.
“two separate scoreboards”
A RevenueCat support engineer says for an AEM campaign, focus on Conversions API and AEM eligibility and use only one SKAN updater, since Meta warns multiple updaters can cause conflicting values and lost signals.
“Meta warns that using multiple updaters can cause conflicting conversion values and lost signals”
Samet Durgun, relayed by Adapty, says 30 to 60% of iOS conversions can be invisible to Meta's attribution, from his experience with no sample. He advises checking against App Store Connect and your backend.
“30-60% of iOS conversions can be invisible to Meta's attribution”
Linkrunner, a measurement vendor, found about two SKAN postbacks per iOS install across about 20 apps in 30 days, against a ceiling of three. It is one vendor's self-selected customers, not Meta-specific.
“The three-postback envelope is a ceiling, not a delivery promise.”
Admiral Media says SKAdNetwork and AdAttributionKit run side by side in 2026. It advises treating missing conversion-value detail as missing rather than zero revenue and comparing cohorts of the same age.
“Missing detail is not a zero-value conversion.”
In the full answer
For readers who sign in- What to do this week3 steps
- The numbers behind it3 figures from cited sources
- The strongest disagreementThe best case against the answer, from a cited source
- How sure we areWhat its confidence rests on, and what would change the answer