iOS Privacy & Aggregated Event Measurement
This is the topic that spawns the most outdated advice on the internet. Let's cut through it: here's what Apple's privacy changes actually did, what's automated now, and the one thing that's genuinely worth your attention.
By the end of this lesson you'll know
- What Apple's ATT changed for advertisers
- What Aggregated Event Measurement (AEM) does today
- Why your reported conversions are partly “modeled”
- The single thing that actually matters now
What Apple did
With iOS 14.5, Apple introduced App Tracking Transparency (ATT) — the pop-up asking “Allow [app] to track you?” A large share of users tap no. That single change cut a big chunk of the signal Meta used to get from iPhone and Safari users, which is a major reason the browser pixel became unreliable.
Meta's answer: AEM
Aggregated Event Measurement (AEM) is Meta's privacy-preserving way to still measure web conversions from opted-out iOS users — in aggregate, rather than person-by-person. It lets your conversion campaigns keep working even though a portion of the audience can't be tracked individually.
Older guides insist you must verify your domain and manually configure and prioritise 8 conversion events. Meta removed that manual step — AEM now aggregates eligible events automatically, and the old configuration tab is gone. If a tutorial is walking you through ranking eight events, it's out of date. (Domain verification is still worth doing for link-ownership and security, just not for this.)
Why your numbers don't match your bank
Because some conversions can't be observed directly, Meta models them — it estimates them statistically from similar cohorts. So a meaningful slice of the conversions shown in Ads Manager are informed estimates, not individually confirmed sales. This is by design, and it's why Meta's reported numbers rarely match your Shopify or CRM totals exactly.
Attribution windows, briefly
An attribution window is how long after seeing or clicking your ad Meta will still credit a conversion to it. The 2026 default is 7-day click + 1-day view (it used to be 28-day click, before iOS forced shorter, more reliable windows). A shorter, click-based window gives you the most conservative, hardest-to-inflate numbers.
What you should actually do
Here's the payoff. You don't fight any of this manually anymore. The whole game is simply: give Meta strong signal and read the numbers sensibly. Concretely:
- Run the pixel + Conversions API together (Lesson 3.4) with good match quality (Lesson 3.3). This is the real fix for lost iOS signal.
- Treat Ads Manager numbers as directional, and sanity-check against your own sales data.
- Pick an attribution window that matches your buying cycle and then compare like-for-like.
Stop chasing one “true” conversion number. In a modeled, privacy-first world, strong first-party signal plus directional reading beats obsessing over perfect attribution every time.
Expecting Ads Manager to match your bank statement to the rupee, or following 2021-era guides that have you manually ranking events. Both lead to wasted hours and wrong conclusions.
Your tracking foundation is done: a dataset installed and verified, meaningful events firing, the Conversions API filling the gaps, and a realistic view of measurement. Now Meta can actually see results — so let's put that data to work by building audiences.
Key takeaways
- ATT let iOS users opt out of tracking, cutting a big share of signal.
- AEM measures opted-out iOS conversions in aggregate — and it's now automatic (no ranking 8 events).
- Reported conversions are partly modeled, so they won't match your CRM exactly.
- The real fix is pixel + CAPI with good match quality, and reading numbers as directional.