Where people leave, and what that actually tells you

Every analytics tool will show you where visits end. Very few owners get anything useful from it, because the highest exit numbers usually belong to pages where leaving is perfectly normal.

Ask for a website review

Exits are only meaningful in context. A high exit rate on a contact confirmation page is success. A high exit rate on your main service page, where people arrive ready to buy, is a leak. Compare pages by intent rather than by raw numbers, and treat the data as a place to start looking rather than as an answer.

Exits are not failures by default

Every visit ends somewhere. If somebody reads your article, gets what they needed, and closes the tab, that is an exit and it is fine. If somebody sends an enquiry and leaves from the thank you page, that is an exit and it is a success.

So the raw ranking of exit pages is nearly useless on its own. What matters is whether leaving from that page, at that moment, means the visitor got what they came for or gave up.

Sort pages by what they were for

Before looking at any numbers, write next to each of your main pages what a visitor is supposed to do there. Then the same data becomes readable, because you know which exits are expected.

A service page exists to produce an enquiry, so exits there are a problem worth investigating. A blog article exists to answer a question and pass the reader onward, so a high exit rate means the onward path is missing rather than that the article failed.

  • Service page: enquiry expected, exits are a warning sign.
  • Pricing page: enquiry expected, exits often mean the cost signal is missing or the next step is unclear.
  • Article: onward click expected, exits mean no next step was offered.
  • Contact confirmation: exits are the desired outcome.
  • Homepage: onward click expected, exits usually mean the message did not land.

The bounce rate confusion

Bounce rate has caused more bad decisions than almost any other number, because a bounce can mean the visitor was disappointed or completely satisfied. Somebody who searched for your phone number, found it, and called you counts identically to somebody who arrived and recoiled.

Modern analytics has moved toward measuring engagement rather than bounces, which helps a little. Even so, treat any single behavioural number as a hint about where to look rather than as a verdict.

What to do once a suspect page is identified

The data tells you where; it never tells you why. To learn why, go and use the page yourself on a phone, in the state a visitor would be in: no prior knowledge, arriving from a search, mildly impatient.

Usually the answer is unglamorous. The page does not say what the thing costs. There is no next step. The evidence is missing. The message is written for people who already know the industry. You will normally see it within a minute, once you know which page to look at.

Look at the entrances too

The other half of the picture is where visits begin. Most owners assume the homepage; in reality, search sends people deep into the site, and those pages are often the ones nobody has reviewed for months.

A page that is a common entrance deserves the same care as the homepage: a clear opening, evidence, and an obvious next step. Many sites have their best introduction on the one page most visitors never see.

Beware conclusions from small numbers

If a page received eleven visits last month, its exit rate is not a finding. Small counts produce dramatic-looking percentages that mean nothing, and building work on them wastes months.

Look for patterns that persist across several months and across pages with reasonable traffic. If the volume is genuinely low everywhere, stop reading percentages entirely and go back to watching real people use the site, which at that scale tells you more.

How Licheo reads this

We look at pages against the job they were given rather than at a league table of exits, and we use the data to decide where to look rather than as the conclusion itself.

Where the numbers are too small to support a conclusion, we say so plainly instead of dressing a guess in a percentage.

Part of a larger guide

This page is one part of Where the money escapes. The other parts:

Questions people ask

What does a high exit rate mean?
It depends entirely on the page. On a confirmation page it means success. On a service page, where people arrive ready to enquire, it means something is stopping them. Judge each page against the job it was given.
Is bounce rate a useful metric?
Only as a hint. A bounce can mean the visitor was satisfied immediately or disappointed immediately, and the two look identical. Treat it as a prompt to go and look at the page rather than as a verdict on it.
Which page should I investigate first?
The one that is both a common entrance and supposed to produce enquiries, usually your main service page. Problems there affect the largest number of people who arrived with intent to buy.
The data shows where people leave, but not why. How do I find out?
Use the page yourself on a phone, arriving from a search, with no prior knowledge. The reason is usually visible within a minute: no cost signal, no proof, no next step, or language written for insiders.
My traffic is too low for any of this. What then?
Stop reading percentages and watch real people instead. At small volumes, three people using your site while you stay silent tells you more than any report, and the findings are usually unambiguous.

Find out where people are giving up

We read the data as a starting point, then go and look at the pages themselves, which is where the actual answer always is.

Ask for a website review Read the full guide