Email Open Rates After Apple Mail Privacy Protection: What 'Opened' Really Means

How is an email open counted?

A newsletter carries a very small image. When the email program downloads that image, the sender's server sees the request and records an open. In MAJ Leads the image address holds a signed token for the campaign and the lead, so a request for it is logged as an open for that one person. Nobody has to click anything, and nobody has to read anything. The open only says that the image was fetched.

What did Apple change with Mail Privacy Protection?

In its announcement dated 7 June 2021, Apple said Mail Privacy Protection masks a person's IP address so it cannot be linked to other online activity or used to find their location, and that it "helps users prevent senders from knowing when they open an email." If the program fetches the image for you, the fetch no longer tells the sender what you did. An open can be recorded when nobody looked at the message.

Apple's announcement does not say how many people use the feature or how a sender should adjust. We do not have a primary source for those, so we do not give a figure for how inflated a typical open rate is. Other things also fetch images without a reader: some company security scanners open every link and image in an inbound message. That is common knowledge among senders, but we cite no study for it.

How does MAJ Leads treat an open?

It counts the open and then asks whether a person probably made it. A request is flagged as automated when it comes from an address that belongs to Apple, when it carries no user agent at all (a real mail program always sends one), or when its user agent matches a list of bots, crawlers and known security scanners. The check is deliberately cautious: it leans towards counting an open as human, so a flagged open is very likely automated, and some automated opens will not be flagged.

The report keeps both numbers. It shows the total with the note that opens may include automated opens, and the automated count beside it. We chose not to hide the doubt inside one clean percentage.

Which numbers should you trust instead?

The sorting is ours, not a standard. It follows from how the numbers are produced. A subject-line test judged on opens alone can reward the version that Apple's service fetched more often, so we would judge it on clicks or replies. Sending to people who agreed to hear from you matters more than any of these numbers; our guide to single and double opt-in covers consent, and the bulk sender rules cover what Gmail, Yahoo and Microsoft expect.

What should a small UAE team do with this?

1. Put one link or one question in every email that matters, so a click or a reply is possible.
2. Compare campaigns by clicks, replies and booked leads. Use opens only to spot a message that was not delivered at all.
3. Read the automated count beside the opens. If it is large, the open rate is telling you less than usual.
4. Do not remove people from a list for not opening. An Apple user may have been counted as opening, and another may have read it with images off.

Where does this leave your newsletters?

Newsletters in MAJ Leads go to your own leads and customers, with an unsubscribe on every message. The report is built to show what is known and what is doubtful. See how it works and pricing. Bounces and unsubscribes are the other numbers worth watching, and hard and soft bounces are covered in the next guide.