App Tracking Transparency (ATT) is Apple's privacy framework for iPhone and iPad. Since iOS 14.5, released in April 2021, apps must show a prompt asking permission before they track a person across other companies' apps and websites or share data with data brokers. If the person chooses Ask App Not to Track, the app can't use Apple's advertising identifier for them and must not track them by other means.
Why ATT changed Meta advertising
Before ATT, Meta's apps could connect what someone did on your website or app with what they saw on Facebook and Instagram, in detail and over long windows. After ATT, for iOS users who opt out, that connection is restricted. The effects for advertisers included:
- Fewer reported conversions from iPhone users, and more conversions that are modelled rather than observed.
- Shorter attribution windows. Meta removed the 28-day click and longer view windows and moved the default to 7-day click and 1-day view. See attribution window.
- Delayed and aggregated reporting for some conversions.
- Smaller retargeting and lookalike pools where the data comes from iOS users who opted out.
- New measurement protocols, including Aggregated Event Measurement for web and app events, and Apple's SKAdNetwork for app installs.
What ATT doesn't do
- It doesn't stop ads being shown to iPhone users.
- It doesn't stop first-party tracking on your own website. Your store still knows what happened on your site.
- It doesn't apply to Android, which has its own separate privacy changes.
How advertisers adapted
- Server-side tracking. The Conversions API sends events from your server, alongside the browser pixel, which makes your data more complete and resilient.
- Better matching. Sending hashed email and phone numbers with events raises event match quality.
- Business-level metrics. Many brands now judge Meta on MER and store data, not only on platform ROAS.
- Experiments. Incrementality tests and lift studies measure impact without relying on individual tracking.
- Broader targeting and stronger creative, since Meta's own modelling does more of the work.
Worked example
Example: a store compares its own sales with what Meta reports.
| Store data | Meta-attributed | |
|---|---|---|
| Purchases in the week | 500 | 380 |
| From iOS users (estimated) | 250 | 160 |
Some of the gap comes from people who opted out of tracking, some from people who found the store another way. You can't split it exactly. That's why total spend against total revenue, through MER, sits next to platform numbers in a sound report.
ATT today
Years later, ATT is simply part of the landscape. Opt-in rates vary by app and region, and Meta's systems have adapted with modelling and server-side data. The practical lesson stays the same: own your first-party data, send it through the Conversions API, and measure the business, not only the platform. Our guide to Meta ads attribution settings covers how to read today's numbers.
Frequently asked questions
What is App Tracking Transparency?
App Tracking Transparency is Apple's rule, introduced with iOS 14.5 in April 2021, that apps must ask permission before tracking people across other companies' apps and websites. If a person declines, apps can't use Apple's advertising identifier for them.
How did iOS 14.5 affect Facebook ads?
Meta receives less data about iPhone users who opt out of tracking, so some conversions are modelled, delayed or missing. Meta shortened attribution windows, introduced Aggregated Event Measurement and leaned more on the Conversions API and modelling.
Does ATT stop my ads showing to iPhone users?
No. Ads still show to iPhone users. ATT limits tracking across apps and websites, which affects measurement, retargeting pools and some optimisation signals, not whether ads can be delivered.
How can I reduce the impact of ATT on my Meta ads?
Set up the Conversions API alongside the pixel, send hashed customer information to improve event match quality, judge results on store-level metrics such as MER, and use incrementality tests for big budget decisions.
