Most email tools track opens with a tracking pixel: a single invisible image embedded in the email, hosted on the sender's server. When the email client loads images, it requests that pixel, and the request gets logged as an open. It's a genuinely old technique, and it worked reasonably well for years. It doesn't work cleanly anymore, because a large share of pixel requests today are fired by a privacy feature, not a person opening the email.
How the pixel actually works
The tracking pixel is typically a 1x1 transparent image with a unique URL per recipient, per send. Nothing about the email's visible content changes. When the client fetches that image to display it, the request hits the sender's server, and that hit gets attributed to the recipient opening the message. Gmail, Outlook, and most major clients set this up largely the same way, whether the underlying email tool is Mailchimp, HubSpot, Gmail's own read receipts, or a custom send.
Why Apple Mail wrecks the number
Apple's Mail Privacy Protection routes images in an email, including the tracking pixel, through Apple's own proxy servers before the message ever reaches the recipient's inbox, and it does this whether or not the person ever opens the email at all. That proxy fetch fires the pixel. The result is that a meaningful share of Apple Mail users show up as having opened every single email sent to them, including ones sitting unread. Since Apple Mail carries a large share of overall email opens, this doesn't just add noise to one segment, it can push a whole campaign's reported open rate up without a single additional person actually reading anything.
What this means for the number in your dashboard
An open rate that includes Apple Mail Privacy Protection traffic is not simply your real open rate plus some margin of error, it can be systematically inflated to the point of being close to meaningless as an engagement signal on its own, particularly if you're comparing it against an older baseline collected before this feature existed on Apple's devices, or against a segment with a different mix of email clients. A jump in reported opens after adding Apple Mail subscribers or after Apple's own adoption grows says more about the proxy than about your subject lines.
What still works
Clicks and purchases are not affected by MPP's pixel proxying the same way. A click requires the recipient to actually tap a link in the email, which the proxy does not do on their behalf, so click-through data stays a meaningfully more honest signal of real engagement than open data now is. This is also where click tracking earns its keep as a replacement metric: instead of asking whether an email was opened, a tagged link inside it asks whether anyone did anything as a result, which is closer to the question most people actually care about anyway.
What to do about it
Keep watching opens for very rough, directional signal, like whether a send obviously failed to deliver, but stop treating open-rate changes as a reliable read on subject-line performance or list health. Weight click-through rate, and whatever happens after the click, more heavily than you used to. If a link in the email is worth tracking on its own, tag it with UTM parameters so it's attributable in your site analytics independent of whatever your email platform reports, and if the destination is a bio-style page rather than a single URL, a tracked link-in-bio page gives you referrer and click data on what happens after the tap, which a pixel embedded in the email itself was never able to tell you regardless of whose privacy proxy got in the way.