Email marketers can't trust open rates inflated by Apple MPP and link scanners

Open rate quietly stopped meaning what everyone still reports it as. Apple Mail Privacy Protection pre-fetches images in email, which registers an open whether or not a person ever looked at the message. A large share of the opens on a normal list now arrive before anyone could plausibly have read anything. The uncomfortable version of that: a rising open rate with flat revenue is not a win, it is measurement drift, and it flatters exactly the campaigns that are not working. Clicks are not clean either. Security scanners on business mailboxes follow links automatically, often seconds after delivery. The tell is a click arriving with no matching open, or several clicks from the same address in the same second. What I ended up trusting instead is revenue per email as the top line, clicks as a directional signal rather than a verdict, and conversions as the one number that settles an argument. Opens I now read only in the negative direction. A collapse is real information. A rise is not. The part I have not solved is reporting it to other people. Nobody wants to hear that the headline metric is unreliable. When someone senior or a client asks how the program is doing, the honest answer is a revenue number plus a caveat about attribution, and the caveat is usually the first thing that gets edited out.

opens have been machine noise since the pixel started getting prefetched at 4am from a server somewhere and everyone just kept putting the same graph on the same slide anyway. last quarter i sat in a review where someone proudly showed opens up 12% and nobody in the room asked a single question, it was december, people bought iphones, thats it. and the 'a few at least understand clicks matter too' part, cool, congrats, thats the floor not the ceiling. clicks are also getting cooked by link scanners now, my own security gateway rewrites every url in a send so half my click data is basically my own IT department clicking links on purpose. i dont even open the open rate column anymore unless a client asks, and when they ask i say we look at trends in aggregate and move on quickly. honestly the whole thing is we sold a metric that a phone can generate for us and now we have to keep pretending it means something

Click rates aren't the clean replacement either, a chunk of those come from Microsoft's link scanners and not humans. You just swapped one fake metric for a smaller fake metric.

Trust it near zero as a performance metric, but don't throw it away either. The reason flat opens with falling clicks looks weird is that the two numbers aren't measuring the same thing anymore. Clicks still require a human; opens mostly don't. MPP and Gmail image caching pre-fetch the pixel regardless of what the recipient does, so the open count tracks list size and sending cadence, not engagement. Stable list + regular sending = flat opens while real attention drops, which is exactly the pattern described here. What I'd report instead: clicks per delivered, and reply rate if the sending supports replies. Both cost a bot more effort to fake, so they degrade more honestly. Any open-based adjustment you apply today is a guess that ages badly, so I stopped presenting adjusted opens entirely. One caveat on dropping opens completely: keep them as a hygiene tripwire, just not a KPI. A sudden open collapse on one domain group (only Microsoft recipients, say) is usually a blocklist or spam-folder problem, and it shows there weeks before clicks would. Alarm, not dashboard number.

Your open rate is holding steady because it’s heavily inflated by automated background activity. Apple’s Mail Privacy Protection (MPP), Gmail image caching, and corporate security filters automatically pre-fetch tracking pixels before a recipient even touches the message, triggering false opens. That’s why your clicks and conversions are telling the real story while opens look artificially flat. Focusing on Click-Through Rate (CTR) is definitely the way to go. I also keep an eye on Click-to-Open Rate (CTOR) to compare relative creative engagement between campaigns, but you have to take it with a grain of salt since the inflated open count warps that denominator too. For stakeholder reporting, I still include open rates so nobody wonders where they went, but I clearly label them as an estimated, directional metric rather than a measure of true engagement.

Since Apple MPP started pre-fetching images, I don't know how to interpret open rates anymore. It's like half of them are counted as soon as the email lands regardless if the lead actually read it. So now I'm not sure if "opened X times" is actually reliable since they might be pre-fetches, and I have no way to tell as my tool still reports it as a high open. Is open rate still important or are you using it with something else/dropped entirely?