How does MPP caching bias engagement data?

In 2021, Apple rolled out Mail Privacy Protection (MPP) in iOS 15 and macOS Monterey. The result: open rate data became much less trustworthy overnight for most email programs.

Here's what MPP does: when an Apple Mail user receives your email, Apple's proxy servers prefetch and cache all the images in it, including your tracking pixel, often before the subscriber even taps on the message. From your ESP's perspective, it looks like an open. Your ESP records it as an open. But the subscriber might never have actually read the email.

For lists where 40-60% of subscribers use Apple Mail, this can inflate open rates dramatically. An apparent open rate of 50% might reflect genuine engagement from only 25-30% of your list. The opened-without-reading subscribers are invisible in your data.

The downstream effects are significant. Re-engagement thresholds set by open data become unreliable. Subscribers who appear active because MPP fires their pixel might not have read your email in months. Suppressions or re-engagement campaigns based on opens alone will miss real inactives.

Better ESPs now flag MPP opens separately. Klaviyo and Mailchimp both report machine opens vs. human opens where they can detect the difference. If your ESP does this, use the filtered view for any engagement-based decisions.

Click data remains relatively reliable. Clicks require an actual human interaction and aren't affected by Apple's prefetching. Weight click behavior over open rates for any meaningful segmentation or deliverability monitoring decisions.