Where you saw it: Open rates jumping to unrealistic levels, a large share of opens from Apple-owned IP ranges, or geographic reporting collapsing into a few proxy locations.
Why it happens
- The feature is opt-in at setup and a large majority of Apple Mail users accepted it, which means most iPhone-read mail is affected.
- Apple fetches the tracking pixel automatically, and your platform counts that fetch as an open.
- The proxy masks the real IP, so location and device reporting from those opens is fiction.
- It applies to anyone reading in Apple Mail, including Gmail addresses read through the Apple Mail app.
The fix, in order
Stop using open rate as your primary metric
It is inflated by an unknown amount that varies with how much of your list uses Apple Mail. Any trend line crossing the feature's rollout is measuring Apple, not your audience.
Move engagement rules onto clicks and replies
Clicks require a human action. If your re-engagement and suppression rules run on opens, they are being driven by a proxy.
Segment Apple Mail opens where your platform reports them
Several platforms flag proxy opens. Excluding them gives a smaller and far more honest number.
Rebuild your benchmarks after the change
Comparing today's open rate against a 2020 baseline compares two different measurements. Reset the baseline rather than celebrating the jump.
Use a click or reply for sunsetting decisions
An address that has not clicked in a year is dormant regardless of how many opens the proxy recorded.
How to know it worked
The short version
- Your engagement segments are built on clicks rather than opens.
- Reported open rate and click rate move independently, which is the sign you are reading a real signal.
- Apple proxy opens are excluded or labeled in your reporting.
Questions people ask
Does this affect all Apple users?
Those who enabled it, which is most of them, and it applies to anyone reading in Apple Mail rather than only to iCloud addresses. A Gmail address read on an iPhone in Apple Mail is affected.
Can I detect a proxy open?
Often. The request comes from Apple's IP ranges and frequently arrives immediately after send. Several platforms flag these, and none of them catch every case.
Is open tracking worthless now?
Not worthless, and no longer a measurement of readership. Treat it as a weak upper bound and put your decisions on clicks and replies.