How Apple's tracking prompt changed what it costs to find a customer

What the prompt actually requires

Apple's rule applies from iOS 14.5, iPadOS 14.5 and tvOS 14.5 onward, released in April 2021. Before an app can access the device's IDFA — a unique identifier used to target and measure ads — or otherwise track a user's activity across other companies' apps and websites for advertising purposes, it must ask through the App Tracking Transparency framework and receive permission, according to Apple's developer guidance.

Apple's definition of tracking is broad. It covers showing someone an ad built on data from other apps, sharing a user's location or email list with a data broker, handing an advertising ID to an ad network for retargeting, and any third-party code library that joins data across apps to target or measure a campaign. Developers cannot withhold app features from anyone who declines, and cannot substitute a hashed email or device fingerprint for the identifier they were denied. Data sharing used solely for fraud prevention, security or credit checks is exempt.

The moment users got a real choice

When the prompt started appearing, most people said no. Research on Apple's policy, summarized by UCLA Anderson, found that roughly 80% to 85% of Apple device owners declined tracking once asked directly.

That mattered because so much of mobile advertising ran on the identifier Apple gated. The IDFA let an advertiser connect what someone did in one app to what they did in another, so campaigns could target people who resembled past buyers and be measured against whether they converted. Losing that identifier for most users meant losing the data trail that targeting and measurement had been built on.

What Meta told its investors

On Meta's fourth-quarter 2021 earnings call in February 2022, chief financial officer Dave Wehner gave the clearest public number on the change's cost. "We believe the impact of iOS overall as a headwind on our business in 2022 is on the order of $10 billion," he told analysts, according to the earnings call transcript.

Wehner was explicit that the figure was a rough estimate rather than a calculation: "We're just estimating what we think is the overall impact of the cumulative iOS changes... We can't be precise on this. It's an estimate." He pointed to two specific mechanisms behind the number: reduced accuracy in ad targeting, which raised the cost of driving a given outcome, and greater difficulty measuring which ads actually led to a conversion.

Why small advertisers felt it hardest

A study of the policy's effect on e-commerce advertisers, forthcoming in Management Science and produced by researchers from Northwestern, Columbia, UCLA Anderson and the University of Maryland's Robert H. Smith School of Business, along with a fifth co-author whose affiliation is given as the World Bank by the Smith School's summary and as Hamburg University by UCLA Anderson's, found the damage concentrated among small firms. Small businesses drew roughly 60% of their traffic from Meta ads, compared with about 30% for larger competitors, according to UCLA Anderson's summary of the research — leaving them less room to fall back on other channels once Meta's targeting weakened.

The two universities' write-ups of the same study describe the size of the hit somewhat differently. UCLA Anderson's summary reports revenue growth at small, Meta-dependent firms down 37% relative to less-exposed competitors. The Smith School's summary describes a 60% revenue drop, relative to less-exposed firms, for small e-commerce firms heavily reliant on Facebook advertising, alongside a 37.1% drop in click-through rates industry-wide. The figures are not directly comparable, but both point to an effect concentrated at the small end of the advertiser base.

The Smith School's write-up quotes one of the study's researchers, Daniel McCarthy, saying affected businesses "weren't able to just rotate their marketing dollars to some other marketing channel" once Meta's targeting weakened — unlike larger brands, which could spread budgets across more platforms.

We can't be precise on this. It's an estimate.

Where the ad dollars moved

Money moved in response to the weaker targeting. The Smith School's summary reports Meta saw a 6.8% decline in ad spending as companies shifted budgets to Google. UCLA Anderson's summary separately reports that Meta's overall online advertising market share fell 4.4% following Apple's April 2021 change, with the study suggesting Google captured much of the displaced spending.

The two figures use different measures — one in ad-spend dollars, the other in market share — and the write-ups do not say whether they describe the same period or the same underlying shift in spending.

The rule today

The framework has kept moving since 2021. Apple's developer documentation states that, starting with iOS 27.2 and iPadOS 27.2, developers in the European Union will have the option to show an alternative consent prompt that includes an "Additional Information" button, and will be able to reprompt a user one year after a previous choice, whether it was to accept or decline — a change specific to the EU.

For a founder or operator, the practical result of the past several years is arithmetic rather than sentiment: a majority of iPhone customers cannot be individually targeted or tracked through Meta's ad platform the way they could before April 2021, and the businesses most dependent on Meta for customer acquisition have carried the largest share of the resulting cost.