Analytics

iOS Privacy Changes: What Actually Broke Tracking

The real impact of ATT on tracking is narrower than the panic suggests — what actually broke, and how CAPI and first-party data recover visibility.

Every iOS privacy update gets treated as another nail in the coffin of tracking, but the real impact is far narrower than the panic suggests — most of what "broke" was fragile to begin with.

ATT killed part of tracking, not all of it

App Tracking Transparency killed device-level cross-app tracking for users who opted out — already a majority of iOS users — but it never touched web tracking, server-side Conversion APIs, or aggregated modeling. Campaigns that relied entirely on last-click pixel attribution were hit the hardest; campaigns already running CAPI alongside the pixel barely noticed a drop in reported conversions.

The fix is not bypassing ATT, but reducing reliance on a single signal

Server-side CAPI plus modeled conversion fallback, which most ad platforms now offer out of the box, restores most of the lost visibility. Collecting first-party data at the landing page level (capturing email before the redirect, not after) provides a signal that survives whatever the OS does next.

The impact varies sharply by GEO

GEOs with a lower iOS share (most of Latin America, Southeast Asia) are far less exposed to this problem than the US/UK/DACH — worth factoring in when prioritizing GEOs, if tracking reliability matters more than raw reach.

In short

  • ATT affected device-level cross-app tracking, not web tracking or CAPI.
  • Campaigns relying solely on pixel and last-click were hit hardest.
  • Server-side CAPI and first-party landing-page data recover most of the visibility.
  • GEOs with a low iOS share are less exposed to the problem — factor that in when choosing markets.

If you're planning to enter this direction — run the economics before launch, not after the first thousand dollars spent.

Let us model your funnel before any spend

Tell us about the offer and the target market — we will come back with a volume estimate, a payout model and a test timeline.