Why open rates are unreliable

Open rates are the most quoted email metric and the least trustworthy. Worth understanding before you make decisions on them.

How an open is measured

By a tiny invisible image in the email. If it loads, we record an open. That is the only mechanism available, and it fails in both directions.

It over-counts

Apple Mail's privacy protection fetches that image for every message, whether or not the recipient looked at it. Several other clients and corporate security scanners do the same.

So a portion of your "opens" are machines. On an audience with many Apple Mail users, open rate can be inflated substantially.

It under-counts

Clients that block images by default record no open even when the message was read carefully. Plain-text readers never register at all.

What to do instead

• Judge campaigns on clicks. A click is a deliberate human action and is measured directly.
• Use opens only as a relative signal between your own campaigns, sent to similar audiences in similar conditions.
• Never use opens to prune a list. "Has not opened in six months" sweeps up people who read everything in a client that blocks images. If you want to identify disengagement, use clicks over a long window.

What a sudden change means

A sharp drop in open rate with clicks unchanged is usually a measurement change, not an audience change — a shift in which clients your recipients use.

A drop in both opens and clicks is real, and usually points to delivery rather than content: check whether the campaign was paused or heavily suppressed — see Why is my campaign paused?.

Next: open Metrics and compare campaigns on click rate instead.