Every email tool reports an open rate. Almost none of them explain that the number is an estimate built on a single fragile assumption: that your reader's software loads remote images. When that assumption holds, the number is useful. When it does not, the number is quietly wrong in both directions.
What is really being counted
An "open" is one HTTP request for a one-pixel image. That is the whole mechanism. It is not a measure of reading, attention, or intent. It is a measure of an image loading, which usually happens because a human opened a message, and sometimes happens for reasons that have nothing to do with a human at all.
The two ways it goes wrong
Undercounting. Corporate mail gateways strip remote images as a security policy. Plenty of individuals block them too. Those readers are invisible to every tracker on the market. If you sell to enterprises, a meaningful share of your most important readers will never register an open.
Overcounting. Since Apple introduced Mail Privacy Protection, Apple Mail prefetches remote images whether or not the message is ever opened. Those prefetches look exactly like reads. If a large part of your list is on Apple Mail, your open rate contains a number of opens that never happened.
So what is it good for?
At the level of a single email to a single person, an open is a strong positive signal and a non-open is close to meaningless. That asymmetry is the whole trick. Use it that way.
- Trust a green tick. Something loaded your image. In one-to-one email that is almost always your reader.
- Ignore a grey tick. It carries no information. Never write "they did not read it" in your CRM.
- Compare against yourself, never against a benchmark. Published averages are computed across wildly different audiences and clients.
- Prefer clicks when the stakes are high. A click is a deliberate human act. An open is a request from software.
A better question
Instead of "what is my open rate", ask "how long did it take this person to open it, and how many times did they come back". Those two numbers survive all of the problems above, because they only ever describe the people who did register, and they describe them precisely.
