How does MPP work? (Image proxying, IP obfuscation)

When someone with Apple Mail and MPP enabled receives your email, here's what happens before they ever see it: Apple's servers fetch every image in the message, including your tracking pixel, and cache that content on Apple's infrastructure. Your ESP interprets that image load as an "open." By the time the person actually looks at the message (or doesn't), your tracking system has already recorded the event. The reader's real IP address, device type, and location are never passed back to your ESP. Apple's proxy IP shows up instead.

The pixel-fetching mechanism is the core of how MPP disrupts open tracking. Traditional open tracking works by embedding a tiny 1x1 transparent image in the email body with a unique URL pointing to your ESP's tracking server. When someone's email client loads the image, the ESP records a hit and marks the message as opened. MPP intercepts this by pre-loading images at Apple's level, before the message reaches the reader's Mail app. The pixel fires whether or not the person opens the email. For many senders, Apple fetches the pixel as soon as the email arrives in the inbox, even if the subscriber never looks at it.

IP obfuscation is the second layer of MPP's impact. Even if you tried to infer location from network requests rather than pixel data, you couldn't: Apple routes all those requests through its own proxy servers, so the IP address your ESP sees belongs to Apple, not your subscriber. Geolocation-based features like send-time optimization by timezone or country-level targeting become unreliable for Apple Mail users. Both your open rate data and your location data for this segment are compromised by the same mechanism.

It's worth knowing what MPP doesn't touch. Click tracking isn't pre-fetched: Apple doesn't auto-click links on subscribers' behalf. Unsubscribe tracking tied to link clicks isn't affected. Conversion tracking that happens on your site after a click isn't affected either. This is exactly why click-through rate has become the more trustworthy engagement signal since iOS 15. Clicks are human-initiated actions. Image loads, for Apple Mail users, are not.

If you want to identify which subscribers are affected, look for contacts showing very high open rates but near-zero click rates. That pattern at scale is a strong indicator you're seeing Apple proxy fetches rather than genuine engagement. Most ESPs now flag Apple Mail proxy opens explicitly in their reporting. Once you've identified that segment, use click activity as your engagement signal for those contacts, and set re-engagement thresholds based on clicks rather than opens. That gives you a signal MPP can't inflate.