Average session duration in Google Analytics

What is average session duration in Google Analytics?

‘Average session duration’ is a metric that measures the average length of sessions on your website. Google Analytics begins counting a session from the moment a user lands on your site, and continues counting until the session ends (i.e. the user exits the site or is inactive for a predetermined amount of time).*

For example, if a user lands on an ecommerce site’s homepage at 12:00, then navigates to a category page, a product page, their shopping cart, and then exits from the checkout page at 12:30, all of those actions (also known as ‘hits’) happen within the same session, which began at 12:00 and ended at 12:30. The session duration would be 30 minutes:

* By default in Google Analytics, a session can last for up to 30 minutes without a user interacting with your pages (for example, by clicking on a call to action or navigating to another page on your site). After 30 minutes, GA will stop counting and the session duration will be ‘0’, and their visit will be counted as a bounce.

Where can you see average session duration in Google Analytics?

The average session duration metric can be found on some of the reports in GA that are related to how users view your website and pages, such as:

• The Audience > Overview report, where Session Duration can be chosen from the metric drop-down menu:

• The Audience > Mobile > Overview report, which shows you average session duration by device category on the report’s data table:

• The Behavior > Site Content > Landing Pages report, which shows you average session duration by landing page:

To get even more specific with your reporting, you can also set session duration as a secondary dimension in some data tables to customize and narrow down your results:

How is average session duration calculated for a website?

A website’s average session duration is calculated by dividing the total duration of all sessions by the number of sessions on a site (over a specified time period):

average session duration = total session duration / total sessions

For example, let’s say your site had 120 sessions in a week. 100 sessions were 30 minutes long, and 20 sessions were 10 minutes long, for a total of 3200 minutes. Your website’s average session duration would be ~27 minutes (3200 total duration / 120 total sessions).

What’s the difference between ‘average session duration’ and ‘time on page’ in GA?

The ‘average session duration’ metric measures the duration of a user’s entire visit to a website, while ‘time on page’ measures the amount of time the user spends on individual pages during a session.

Using the example from above, the session duration here would be 30 minutes, but the time spent on each page varies from 5 to 10 minutes:

What average session duration does and doesn’t tell you [4 important things to consider]

Session duration helps you understand how long users stay on your website on average, which might be a good starting point for identifying issues in the customer journey.

For example, by referencing the average session duration and time on page metrics together, you can get an idea of how long it might take someone to get through a sales funnel, like homepage > category page > product page > shopping cart > checkout. If it takes users an unrealistically short or unusually long amount of time to get through the funnel, there might be something that needs investigating—but you can’t know for sure without looking past the numbers (more on this later).

Now let’s dig into what average session duration doesn’t tell you, and go over some things to consider when you look at the metric’s data.

Here are four reasons average session duration could be considered a controversial metric:

1. The average session duration metric raises more questions than it answers

Session duration tells you very little about what’s actually happening on your site—in fact, using the sales funnel example above, it only raises more questions, like:

• How did that user get through the sales funnel so fast? Did they miss any important page elements or details?

• Why did it take that user so long to get through the sales funnel? Was there information missing?, did they encounter a website bug?, or were next steps not clear?

2. Session duration doesn’t include exits or bounces

GA doesn’t account for the time spent on the last page of a user’s session—no matter how long they were there. This means that exit and bounce pages are excluded from the average session duration metric because by default, an exit page’s time on page is always ‘0’. This is particularly unhelpful when you want to investigate high-exit or high-bounce pages.

In the bounce example below, the user’s full session happened on one page (which is why bounces are also known as single-page sessions). Technically, the user could have spent a full 30 minutes reading, window shopping, or trying to find their way around your site. No matter how long they were on the homepage in reality though, because they didn’t trigger another request to the GA server, their session duration will be counted as ‘0’.

Here are just three scenarios that could have taken place in the time that wasn’t accounted for:

1. The user could have spent 5 minutes trying to get a broken link to work

2. They could have spent 15 minutes reading and scrolling, and then tried clicking on an element that they expected to take them to another page (but it didn’t)

3. They could have spent 30 minutes consuming the entire page from top to bottom and were satisfied with what they found, so they had no reason to stick around

3. Session duration is not precise

As we mentioned earlier, because GA doesn’t account for time spent on exit pages, average session duration isn’t entirely accurate. Now let’s take a look at why that matters:

Relying on the metric could be problematic if, for example, you want to track the average session duration of visitors who don’t convert.

Average session duration might tell you that your non-converting users spend about 20 minutes on your site overall before they exit—but it doesn’t tell you how much time they spend on the page they exited from. So you don’t really know how long they were on your site altogether before they decided not to convert.

Like the points above, the metric leaves you guessing, with questions like:

• Why are these users leaving without converting? How long did they spend on the last page?

• Did they browse for a while but couldn’t find what they were looking for?

• Is there something on the site—a broken or missing element—that needs to be fixed?

4. Average session duration can be ‘fixed’ or altered

As the site owner, you could potentially ‘fix’ or alter average session duration in two ways:

1. The same way you could potentially trick bounce rates: by using ‘event tracking’ in GA, you can tell the Analytics server to track specific interactions. For example, interactions like embedded video plays or outbound clicks (which take the user to another URL) wouldn’t typically count as events, but when you tell GA to track these actions, you can essentially track (at least some of) users’ time on an exit page, which will affect your average session duration.

2. Changing your site’s default session duration (from between one minute to four hours): as we mentioned, by default a ‘session’ can last for up to 30 minutes without a user interacting with your pages. After 30 minutes of inactivity, GA will stop counting the session. But let’s say you have a lot of content and expect visitors to spend a certain amount of time on a page before interacting (that is, before they trigger another action to the GA server): you could lengthen your default session duration so GA doesn’t stop counting users’ long, ‘inactive’ visits for up to four hours, which will affect your site’s average session duration.

To sum up:

Knowing your site’s average session duration isn’t very useful, out of context. The metric tells you very little: just how long users stay on your site, on average (excluding the page they exit from, which might be what you really want to see if your goal is to optimize your site and improve conversions).

So, what can you do with the metric, then? How can you look at average session duration in context to find out what’s really happening on your site? To do that, you’ll need to use some behavior analytics tools to complement GA and fill in the blanks. Keep reading to find out which tools can help.

2 tools to discover what’s really happening on your site

1. Recordings: see how real visitors interact with your pages (throughout their whole session)

You can use session recordings to go beyond the session duration metric and see how users behave on your site throughout their whole session—from page to page, up to the point of exit. Session recordings give you a clearer understanding of how people click, interact, and move around on your pages so you can identify pain points or blockers they might experience, then optimize as needed.

Session recordings let you watch replays of real, anonymized users moving around and clicking on pages throughout your website so you can get a sense of if/where they:

• Encounter broken links, missing elements, website bugs, or other issues

• Exhibit behaviors like u-turns or rage clicks (which occur when someone repeatedly clicks on the same element)

• Get distracted by pop-ups or other elements and become inactive on a page

2. On-site surveys: find out what your visitors are looking for

On-site surveys will help you find out what real visitors are looking for on your site. Since exit pages aren’t counted in the session duration metric, placing an on-site survey on high-exit or high-bounce pages can be particularly helpful: you can ask users why they leave a page, and hear directly from them (in their own words) how to improve your site.

Place an on-site survey on your high-exit or -bounce page(s) and ask your visitors a specific, open-ended question, like:

• What’s stopping you from buying today?

• What’s missing from this page?

• What are you looking for that you can’t find on this page?