Email click-through rate:
formula and worked example

Why clicks became the metric that matters

Email click-through rate measures the share of delivered messages that produced at least one click. It is the closest direct signal of whether the content and the call to action did their job, and unlike an open it requires a deliberate action from the reader.

Open rates have become less trustworthy because some mail apps pre-load images and register an open without anyone reading. Clicks are not affected in the same way, which is why many marketers have shifted their attention to this figure. It still has to be defined carefully, because there are several versions of it.

Click-through rate also pairs naturally with the next step in the funnel. A click is only the start; the page it lands on, the speed of that page on a phone and the match between the email's promise and the page's content decide whether the click becomes anything. Reporting clicks alongside the sign-ups or purchases they generated stops the metric from becoming a vanity figure.

Formula and the choices hidden in it

• Clicks: ideally unique clickers, so one person clicking four links counts once. Total clicks give a different, larger figure.
• Delivered: emails sent minus bounces. Using sent instead of delivered penalises you for dead addresses.

A related measure, click-to-open rate, divides clicks by opens instead. It tells you how persuasive the content was among people who looked, but it depends on open data that may be unreliable.

Worked example: 420 clicks on 10,000 delivered

A newsletter is sent to 10,200 addresses; 200 bounce, leaving 10,000 delivered. The email platform reports 420 unique clicks.

1. 1Take the emails sent and remove bounces to get delivered: 10,200 − 200 = 10,000.
2. 2Take unique clicks: 420.
3. 3Divide: 420 ÷ 10,000 = 0.042.
4. 4Multiply by 100: the click-through rate is 4.2%.

If 2,000 of those recipients had opened the message, click-to-open would be 420 ÷ 2,000 = 21%, a different number answering a different question.

Interpreting 4.2% without false precision

Published averages are rarely comparable, because they mix lists, industries and definitions. Use the following as conventions to test against your own history.

• Promotional blasts to a broad list normally earn a lower rate than a targeted message to a small, engaged segment.
• Transactional and onboarding emails often outperform newsletters because the reader expects them.
• A falling rate with a stable list usually points to content fatigue; a falling rate with a fast-growing list may be dilution by newer, less engaged addresses.
• Over-long emails with several competing links tend to scatter clicks and lower the unique rate.

List health shapes the number more than most senders expect. Addresses that never engage drag the rate down while still counting as delivered. Many senders find that pruning long-dormant addresses, or moving them to a lower-frequency programme, raises the percentage without changing a single word of content, and also protects their sending reputation.

What a ten percent lift in clicks means

A 10% gain in clicks moves the rate from 4.2% to 4.62%, an addition of only 0.42 points. Because the base is large, rate changes look small while the extra visitors are real: 42 more people arrived on your page. This is why rate and volume should be read together, particularly when comparing sends of very different sizes.

Where email click rates go wrong

• Dividing by emails sent rather than delivered, so list hygiene problems are blamed on the content.
• Comparing unique-click rates from one platform with total-click rates from another.
• Counting unsubscribe or view-in-browser links as clicks on the call to action.
• Comparing two sends made to differently sized or differently aged lists and attributing the gap to the subject line.

Be wary of reading too much into a single send. Day of week, holiday timing, subject line, offer and list freshness all move the rate together, so one strong or weak result proves little. Looking at the median across the last eight to ten comparable sends gives a baseline that is hard to fool, and a new send is then meaningful only if it falls clearly outside the range you usually see.

A simple test you can run this month

Split a list of 10,000 randomly into two halves of 5,000. Send version A with one clear button and version B with three links. If A gets 250 unique clicks (5.0%) and B gets 200 (4.0%), a gap of a full point on 5,000 recipients each is large enough to take seriously; a gap of 0.1 point would be noise. Only change one element per test so you know what caused the difference.

Remember that timing, device and list segment all interact with content. A message that performs at 3% on the full list may reach 7% among customers who bought in the last sixty days. Rather than hunting for a universal best practice, record the rate by segment for a few sends and let your own list tell you where attention pays off.