What is App Tracking Transparency?

App Tracking Transparency, usually shortened to ATT, is the pop-up on iPhones and iPads that asks whether an app may track you. Apple's user privacy page for developers states that in iOS 14.5, iPadOS 14.5 and tvOS 14.5 or later, apps need the user's permission through ATT to track them or access the device's advertising identifier.

Apple defines tracking as linking data from your app with data from other companies' apps, websites or offline sources for targeted advertising or ad measurement. If a person doesn't give permission, Apple says the advertising identifier comes back as all zeros. In plain terms: the ID that let ad platforms follow one phone from an ad to a purchase in another app simply disappears.

What changed for advertisers

Before ATT, a platform like Meta could see that a specific phone tapped an ad and later bought something, then report that sale as its conversion. After ATT, for everyone who declines, that link is gone. Platforms responded by estimating, or modeling, the conversions they can no longer observe, and by relying more on Apple's own privacy-preserving attribution systems.

Apple's systems, SKAdNetwork and the newer AdAttributionKit, report results in aggregate without identifying the person. Apple's documentation explains a few of the rules that shape those reports:

  • Only one ad can win credit for a conversion.
  • Click-through ads always take precedence over view-through ads, the ones that were only seen. Our guide to view-through conversions explains why that matters.
  • The system considers a maximum of six ad impressions for any conversion.
  • AdAttributionKit works with both the App Store and alternative app marketplaces; SKAdNetwork works only with the App Store.
What an 8% to 40% revenue decline means for a $100,000-a-month seller
  • Before ATT$100,000
  • Low-end estimate (8% less)$92,000
  • High-end estimate (40% less)$60,000

A worked calculation applying the 8% to 40% revenue-decline range from Aridor et al. to an example business with $100,000 in monthly revenue; it is not a measured result.

Source: Worked example

What the research found

The most detailed independent study so far comes from Guy Aridor, Yeon-Koo Che, Brett Hollenbeck, Maximilian Kaiser and Daniel McCarthy, forthcoming in the journal Management Science and summarized by Northwestern's Kellogg School. They found that conversion-optimized Meta ads, the type affected most by ATT, saw a 37% reduction in click-through rates after ATT.

Using several methods, they estimated revenue decreases of 8% to 40% for firms that relied most on Meta, compared with less exposed firms. Many of those businesses moved some spending to Google, but the losses still fell mainly on smaller e-commerce and direct-to-consumer brands. The chart below shows what that range would mean for a business doing $100,000 a month.

Why your iPhone numbers deserve a second look

Apple's case for ATT was that tracking had become everywhere. Its 2021 paper A Day in the Life of Your Data said the average app has 6 trackers. Taking that data away made ads less precise, and it also made reports less certain.

Today, a share of the iPhone conversions a platform shows you are modeled estimates rather than observed sales. Models can be good, but they can also drift high or low, and you can't see inside them. A practical warning sign: your iPhone-heavy campaigns report steady sales while your store shows a dip, or the reverse. That usually means the model is filling gaps with assumptions, so treat the number as a claim to check. That uncertainty adds to the reasons platforms often claim more sales than you made.

How to measure well after ATT

You can't undo ATT, and you shouldn't try to get around it. Apple's rules forbid tracking without permission. What you can do is lean on data you own. Matrix ROI, developed by Smith App Studio, puts each platform's claimed conversions next to your verified Stripe or store sales, but the habits below work with a spreadsheet too:

  • Make your payment and order records the scorecard. Our guide to first-party data covers the basics.
  • Compare platform-reported conversions with verified sales weekly, and watch the gap over time rather than any single day.
  • Judge channels by blended results, total revenue against total spend, alongside each platform's own claims.
  • Run simple on-off tests in a few regions before making big budget cuts or increases.