What a good click through rate is, and why the published numbers disagree

What is a good click through rate for email?#

As a working figure, one to two percent of delivered mail is ordinary for a promotional send, and a newsletter that people read closely can sit above it without doing anything unusual. A rate under half a percent on a list that opened well is a message that got read and then not acted on, which is a different problem from the one people usually suspect.

That is an interpretation, and it is worth labelling as one, because the measured numbers on this page are the vendor figures below. What those figures cannot do is tell you whether your own send was good, because they are averages of other people's lists, mailed at other frequencies, to other kinds of readers, counted with other rules. The number to aim at is your own trailing median, and the published figures are for noticing when yours has moved.

What does a click through rate count?#

Before comparing anything, settle three decisions, because two senders using the same phrase can be two points apart purely from these.

• What counts as a click. Mailchimp states that its figure is determined by whether someone clicked on an image, a hyperlink, or a call to action included in the email, which means an image click is a click, and a copy-paste of a URL typed into a browser is not a click.
• Who is counted once. Counting hits and counting people are different numbers. One reader clicking four links is four events or one person, and a report that does not say which is a report you cannot compare.
• What sits underneath the percentage. A rate built on delivered mail is smaller than the same activity built on opened mail, which is smaller again than a rate built on everyone the campaign was addressed to.

The two ratios that come up most are click rate and click-to-open rate, and they answer different questions. A click rate divides clicks by delivered mail and describes the campaign. A click-to-open rate divides clicks by opens and describes the email itself, because it only asks what happened to people who showed up.

What numbers do email tools publish for clicks?#

Both of these come from the vendor's own page, on the same day, describing its own customers' sending.

Now put the two companies side by side on a second metric and the point becomes unavoidable. Mailchimp tells its readers to aim at an open rate of about 34.23 percent. MailerLite reports a median open rate of 43.46 percent. Nine percentage points apart, both from real data, both describing email marketing, neither of them wrong about their own customers. Anyone who quotes a single industry number as a target is quoting one company's base and pretending it is the market.

Why do the two benchmarks disagree by so much?#

Four reasons, and none of them is that somebody lied.

1. Different customer bases. A tool used mostly by online shops and a tool used mostly by newsletters will publish different numbers no matter what it measures, because the audiences behave differently.
2. Different denominators. Delivered, sent, addressed and opened are four plausible denominators and each produces its own percentage.
3. Different counting rules. Unique clickers versus total click events, image clicks versus link clicks, and whether a click inside a forwarded copy counts all move the figure.
4. Different mix of mail types. A dataset that includes transactional receipts and automated confirmations has different traffic in it than one that counts only campaigns.

You cannot fix this by averaging two vendors. You fix it by choosing one denominator, using it on every send, and comparing each new campaign with your own history.

Can I trust a rate that depends on opens?#

Carefully, and the vendors themselves say so. Mailchimp notes on its own benchmarks page that open rates may be affected by Apple's Mail Privacy Protection and that this should be considered when interpreting the data. MailerLite goes further and states plainly that Apple Mail Privacy Protection marks emails opened to Apple clients automatically, which means the real open rate is lower than what you see in your dashboard and in its own article.

That admission has a second consequence people miss. An inflated denominator makes click-to-open look worse than it is, so MailerLite says the actual click-to-open rate is higher than the reported one, and draws the conclusion that click rate is currently the most accurate indicator of newsletter engagement, because it does not depend on an open being recorded at all.

The practical rule: treat any ratio with opens in it as soft, and treat clicks on delivered mail as the harder number. If a change in click-to-open is your evidence, ask whether the opens moved rather than the clicks.

Which ratio tells me which problem?#

A benchmark cannot diagnose any of these, because each pattern is a relationship between two of your own numbers rather than a level. That is the real reason industry figures are weak evidence: they describe a point on a curve nobody shares.

How do I work out my own click rate baseline?#

1. Pick one denominator, and write the choice down where you will still be able to find it in a year.
2. List the last six to twelve campaigns with that rate beside each, plus who was mailed and what the send was for.
3. Take the middle value, not the mean, so one giveaway or one re-engagement send does not set your target.
4. Group by send type if you run more than one, because a receipt, a monthly issue and a sale notice are three different experiments sharing one dashboard.
5. Judge each new send against that median and against the open rate beside it, never against a chart from another company.

How do I get more clicks without tricks?#

The methods the vendors themselves list are unglamorous. Mailchimp's own improvement notes are to create visually appealing and mobile-friendly content, to make the content more dynamic and compelling, and to write clearer calls to action. That is a real list, and it is also the list every designer gives. The three that decide whether a small sender moves are these.

• One reason to click per email. Two offers that both matter is one email fighting itself, and three is a menu nobody reads on a phone.
• Put the reason near the top. The part of a newsletter that gets read is the part above the fold on a small screen, and a link at the eighth paragraph is a link most readers never reach.
• Make the link say what happens. A call to action that reads as a button shape rather than a promise gets fewer clicks because the reader cannot price the trip.

What does not raise clicks is a subject line that oversells. It raises opens, then lowers the click rate, and can push the complaint count up in the same send, which is the one movement that costs you standing with the mailbox providers rather than only your own report.

How are clicks counted in a campaign report?#

Two honest numbers exist and most dashboards show one. The event count is how many clicks happened. The person count is how many readers clicked at least once, and it is the one worth comparing between campaigns, because one enthusiastic reader should not look like five people interested.

In Campaign Kicker, the click figure on a report counts each person once, and a click is only recorded when that address had already been accepted for delivery, which is why a report does not gain clicks from a message nobody received. The tile shows that number as a share of the people who opened rather than of everyone mailed, which is a click-to-open reading, so hold it against your own previous sends and not against Mailchimp's or MailerLite's click rate.

Does a low click rate mean the email was bad?#

Not on its own. A newsletter that informs may be working while producing few clicks, and a click is a poor measure of a reader who got what they needed and closed the message. What a low rate does justify is a look at whether the email contained a single thing worth acting on, and at whether the audience it went to was picked for that thing.

The number that cannot be dismissed this way is the complaint count. Engagement you can explain is a modelling question. Complaints are an answer somebody gave, and they are the measurement the receiving providers act on.

Does an automated email get clicked more than a campaign?#

Usually, and for a reason that has nothing to do with writing. A receipt, a shipping update or a post-purchase follow-up arrives because the reader did something, so the mail is expected and the link inside it answers a question they still have. Mailchimp says as much in its own benchmarks resource, where it notes that high open rates may be found in automated transactional messages such as birthday deals and post-purchase follow-ups, and that those work well alongside marketing automation.

Two consequences for how you read a dashboard. If automated mail is mixed into the same statistics as campaigns, the average flatters the newsletter and the newsletter looks worse than it is. And a comparison between a receipt and a monthly issue is not a comparison, because one is a reply and the other is an invitation. Keep the two counted separately, and set a separate baseline for each.

Does the device a reader uses change the click count?#

It changes both halves of the measurement. An email client that blocks images by default will not fetch the tracking pixel, so the open goes unrecorded while a text link clicked in the same message is recorded normally, which pushes click-to-open above one hundred percent in a report and makes the numbers look broken rather than different. A preview pane on a phone, where a reader taps a link without the client ever loading remote content, produces the same effect.

There is a second, quieter effect. A touch interaction on a small target inside a dense design is easy to miss and easy to mis-tap, which is why the vendors' own improvement lists keep returning to mobile-friendly layouts. If your clicks are low and your opens are low in the same way across every send, look at whether the thing you want tapped is large enough to tap in one thumb.

Should I tag links to see where clicks went?#

Yes, if the tag survives the click and you read the result. Campaign parameters on a link tell the analytics tool which campaign produced a visit, which is the one thing an email report cannot tell you: what happened after the click. A rate tells you people left the email. A tag tells you whether they bought, booked or bounced off the page.

Two limits to respect. A shortened or redirected link can drop the parameters, and a destination that strips them loses the answer. And if you tag, use the same spellings every send, because a campaign named with three variations gives you three small rows and no trend.

What should I change first if clicks are low?#

In this order, because each step is cheaper than the last. Check that the email contained one thing worth doing. Check that it was visible without scrolling. Check that the audience for that send was the audience that wanted it, rather than everyone with an address. Then look at the writing, which is the step most people start from and the one that rarely explains a two-point gap.

If the last two steps are where the problem lives, the fix is a list question rather than a design question, and the pages worth reading next are how to segment an email list and how to clean an email list. A click rate cannot be raised indefinitely by better buttons, and the ceiling is usually the number of people who had a reason to be there.

Common questions