Email metrics: which ones actually matter

The three that decide

Click rate. The share of recipients who clicked a link. It is the only metric that measures a deliberate action — therefore the only one that holds up against measurement artefacts. If you track just one number, track this one.

Conversion rate. The share of recipients who did what you wanted: bought, booked, replied, signed up. You won't find it in your email platform — it lives in your site analytics, through UTM parameters on your links. It is the number that connects a send to revenue.

Complaint rate. The share of people who clicked 'report as spam.' Google requires staying under 0.3 % and recommends less than 0.10 %. It is the only metric with a published threshold, and the only one that can damage all your future sends.

The three that warn you

Bounce rate tells you whether your list is ageing. A rise from one send to the next signals addresses dying off, or a badly cleaned import.

Unsubscribe rate tells you whether your cadence or your content has drifted. A spike on one particular send points at that send — which is rare and precise information.

Delivery rate measures what servers accepted. Useful for list hygiene, but it says nothing about which folder the message landed in: a send can show 99 % delivery and be read by no one.

The two that mislead

Open rate. It is measured by an invisible tracking image loaded on open. Since 2021, Apple Mail's privacy protection preloads those images whether or not the message is read. A share of counted opens corresponds to no actual reading, and another share of real reads is never counted.

What remains: a benchmark for internal comparison. Your Tuesday subject line against your Thursday subject line, measured by the same tool, in the same period. Nothing more.

Click-to-open rate. It divides clicks by opens, so it inherits the problem above — squared. It remains useful for judging content independently of the subject line, as long as you never compare it to click rate.

The only benchmark that helps

Industry averages are everywhere. They almost always come from websites earning a referral fee on the platforms they recommend. They mix unrelated sectors, and they rest on open rates we just showed you cannot trust.

The only benchmark that serves you is yourself, last month. An industry average will not tell you whether your newsletter is improving.

What to check after every send

1. Bounces: more than usual?

2. Complaints: even one per thousand is already high.

3. Clicks, link by link: which one worked, which was ignored.

4. Unsubscribes: a spike, or the usual background noise?

The rest can wait until end of month.