Engaged Means Clicked: Why Open Rate Is a Dead Segment Signal

Apple Mail Privacy Protection has made open rate a measure of how many inboxes auto-load images, not how many people read your email. A segment built on opens is a segment built on a broken signal.

Open rate used to be the standard measure of email engagement, and it is now a dead signal. Apple Mail Privacy Protection, launched in 2021, loads tracking pixels automatically and remotely, so an open is recorded whether or not the recipient ever looked at the message. The result is an open rate that measures how many inboxes auto-load images, not how many people read the email. A segment defined as opened in the last 30 days now includes everyone on Apple Mail, which is a large share of most lists, regardless of whether they care. The signal that still works is the click, because a click requires the recipient to act. An engaged segment should be defined as clicked in the last 90 days, or purchased in the last 180 days, with opens used only as a secondary signal for the minority of inboxes that still report them honestly.

What Apple broke, and what survived

Apple Mail Privacy Protection hides the recipient's IP address and loads remote content privately, which means the tracking pixel fires when Apple loads it, not when the human opens the message. The open rate for Apple Mail users jumped to near 100 percent overnight, and the blended open rate for most lists rose with it, not because more people were reading, but because more pixels were firing. The segment built on opens became a segment of Apple users plus the genuinely engaged, and the campaign sent to it performed worse every quarter while the open rate looked healthy.

What survived is everything that requires an action: clicks, purchases, replies, and unsubscribes. A click is a deliberate act, and it cannot be faked by a privacy proxy. A purchase is the strongest signal of all, and it is the one that should anchor the engaged definition. The rule is simple: define engagement by what the recipient did, not by what their inbox did on their behalf.

Related reading on this site: SegmentMuse and Winback Segments Decay Faster Than Acquisition Segments, So They Need Shorter Clocks cover the neighbouring problems in more depth.

Rebuild the engaged segment on clicks

The new definition is straightforward: engaged means clicked at least once in the last 90 days, or purchased in the last 180 days. The 90-day click window is long enough to include the monthly buyer and short enough to exclude the person who clicked once in spring. The 180-day purchase window catches the seasonal buyer who clicks rarely but buys reliably. Anyone outside both windows is not engaged, whatever their open rate says, and they should not receive the frequent campaigns.

The transition needs a warning label. The day the segment is rebuilt, the engaged list will shrink, often by half or more, and the team will panic. The shrink is the point: the list was never that large, and the campaigns were going to people who were not reading them. The send volume falls, the deliverability rises, because fewer unengaged recipients means fewer spam complaints, and the revenue per send rises with it. Measure the transition by revenue per send and by inbox placement, not by list size.

What to do with the people who only open

The open-only group is not dead, it is unmeasured. Move them to a re-engagement flow, not to the main campaigns: a short series that asks them to click to stay, with a clear choice and a real incentive, and suppress anyone who does not click within 30 days. The suppression is the point. A list that keeps mailing people who never click is a list that trains inbox providers to treat its mail as unwanted, and the deliverability cost lands on the engaged segment too.

The same logic applies to the sunset policy. A subscriber who has not clicked in 180 days and has not purchased in a year should be suppressed from frequent campaigns, not deleted, because deletion destroys the record and the reactivation option. Suppression keeps the door open: a win-back campaign once a quarter, and a return to the engaged segment the moment they click. The segment is a rule, not a list, and the rule is written in clicks.

Frequently asked questions

Why is open rate no longer a reliable engagement signal?

Apple Mail Privacy Protection loads tracking pixels automatically and remotely, so an open is recorded whether or not the recipient looked at the message. Open rate now measures how many inboxes auto-load images, not how many people read the email.

How should an engaged segment be defined now?

By actions: clicked at least once in the last 90 days, or purchased in the last 180 days. A click requires the recipient to act, and a purchase is the strongest signal of all. Opens should be used only as a secondary signal.

Will rebuilding the engaged segment shrink the list?

Yes, often by half or more, and that is the point. The list was never that large. Send volume falls, deliverability rises because fewer unengaged recipients means fewer spam complaints, and revenue per send rises with it.

What should happen to subscribers who only open and never click?

Move them to a re-engagement flow that asks them to click to stay, and suppress anyone who does not click within 30 days. Suppression keeps the door open for a quarterly win-back, while protecting deliverability for the engaged segment.