If your open rate climbed while clicks and orders stayed where they were, the extra opens probably aren't readers. Apple Mail can load an email, including the invisible image that counts as an "open", without the person looking at it. So stop judging emails by opens: measure clicks, replies and orders per send, and fix any automation that waits for someone to open an email before it does the next thing.
The mismatch shows up in a few ways. An open rate in the 50s or 70s next to a click rate of 1 or 2%. A subject line test where both versions "win" with nearly identical, very high opens. A sunset automation that hardly catches anyone. All three come from the same cause.
Why opens stopped meaning "read"
Your email platform counts an open when a tiny image in the email is downloaded. Apple's Mail Privacy Protection changes when that happens: with Protect Mail Activity on, Apple says Mail downloads remote content in the background by default, regardless of whether you engage with the email. The image loads, the platform records an open.
A few details matter when you look at your own numbers:
- It's the Apple Mail app, not every iPhone. Brevo's help on Apple Mail Privacy Protection notes that it affects Apple Mail, not other mail apps on Apple devices such as Gmail. A list full of Gmail-app users is affected less.
- Your platform may have changed how it counts. The same page says Brevo's open rate has included Apple Mail Privacy Protection opens since 6 February 2025, and bot activity since 8 July 2025, and that you can filter both out. A sudden jump in your reports can come from a counting change like this rather than from your readers.
- Clicks can be inflated too, but less. Security software that checks links can register clicks. Brevo lists bot activity as a filter for this reason. Clicks are still a much better signal than opens; just don't treat a click within seconds of delivery from many addresses as real interest.
What to measure instead
| Instead of | Track | How to read it |
|---|---|---|
| Open rate | Click rate: unique clicks ÷ delivered | Compare each send with your own last ten, not with industry averages |
| "Opened in the last 90 days" as engaged | Clicked or ordered in the last 90–180 days | A smaller group, but the people in it did something |
| Subject line test on opens | Same test, winner by clicks or orders | Needs a bigger list to show a difference; skip it if your list is small |
| Opens as a sign an email arrived | Replies, clicks, orders, and your platform's bounce and complaint numbers | A drop in clicks across several sends is the early warning now |
| Revenue "from opens" | Orders placed after a click on the email | Use your store's or platform's attribution window and keep it the same between sends |
Replies are worth more than they look. If your emails come from a person and invite answers ("reply and tell me what you're making with it"), each reply is a reader you can be sure of.
You can still look at opens for people who don't use Apple Mail, if your platform separates them. Brevo, for example, shows the open rate with Apple MPP opens included or excluded. That number is fine for spotting a sudden drop. Just don't make decisions on the blended figure.
Fix automations that wait for an open
This is where inflated opens do real damage, because nobody sees it happen. Go through every automation and look for steps like these. The middle column assumes your platform counts Apple Mail's automatic opens as real ones:
| Step | What goes wrong | Change it to |
|---|---|---|
| "If opened email 1, send email 2a; else send 2b" | Apple Mail users go down the "opened" branch whether they read it or not | Branch on "clicked any link in email 1" or "placed an order since" |
| "Resend to non-openers after 3 days" | Apple Mail users don't get the resend, even if they missed the first one | Resend to non-clickers, once, with a new subject, or drop the resend |
| "Wait until opened, then…" | The wait ends at delivery for Apple Mail users | Wait a fixed time, then check for a click or order |
| "Not opened in 90 days" starts a sunset flow | Apple Mail users don't qualify and stay on the list indefinitely | Use "no click and no order in 180 days", as in the re-engagement sequence |
| Segment "opened in last 30 days" for big campaigns | The segment includes everyone on Apple Mail | Use "clicked in the last 90 days or ordered in the last 180" |
Platforms handle these steps differently, so check yours. Brevo says that for Apple MPP users an "Email opened" trigger, a conditional split based on opens, and a "wait until" step based on an open will not work, and that its open-based segments leave Apple MPP opens out by default unless you change that. Another platform may count those same opens as real. Read your platform's own help page on Apple Mail Privacy Protection before deciding which branch a person ends up in.
Where this leaves your reporting
For a weekly or monthly report, three numbers per campaign are enough: clicks, orders (or replies, if you don't sell online) and unsubscribes plus complaints. Write them next to the same numbers for your last ten sends. When clicks fall for several sends in a row, look at whether your emails are landing in spam before you blame the subject lines.