Average session duration

Useful for: Providing an indication of how ‘sticky’ a website is, and whether any CRO work has yielded improvements.

Be careful of: How this metric is reported in Google Analytics – it’s not the most accurate!

Our preferred tool: Google Analytics (albeit with some tweaking).

On the face of it, the average session duration metric from Google Analytics sounds great. After all, as SEO’s we all know that the better and ‘stickier’ a website is the easier it is to rank. However, not all is as it seems!

We’re just going to throw this out there, but Google doesn’t measure the time a user spends on the last page of their visit to your site. This is because (by default at least) Google Analytics only fires off a tracking pixel on a unique page load, and not when a user exits away from a site. The website Jimmy Data has a really good explanation of this, told through a real life analogy, so be sure to visit their website if you want more information on how Google calculates this metric.

The main takeaway here is that average session duration is only calculated from a sample of your total users, and the higher your bounce rates are, the more inaccurate this metric will be for you.

All is not lost however, and there are third party tools which can more accurately measure average session durations. Our recommendations are Hotjar and Mixpanel, both of which are very good conversion rate optimisation (CRO) analytics suites.

Another option to improve Google Analytics’ average session duration is to adapt your tracking code to fire off a non-interaction event every few seconds, thus capturing those visitors who either bounced or spent a long time visiting the final page on their visit to your site. Littledata have written a script for this that will fire off a non-interaction event every 10 seconds.

Even if you decide not to adapt the default Google Analytics settings, the key takeaway here is that yes – average session duration is a good indicator of how sticky a website is, but it should not be considered a reliable KPI for reporting purposes. If you are going to include this metric in your reporting, we’d strongly recommend including a disclaimer so that stakeholders are aware of its shortcomings.