August 4, 2020 by Adrian Durow . Analytics
It’s often not spammy traffic. And it’s definitely not accurate. How can a user genuinely spend 0 seconds on a site or a page? They can’t.
As highlighted in the previous post, the duration of a session is the time it takes from the first hit (which is typically a page view) sent to analytics in a session, to the last hit to be sent. So in the below example, the session duration is 0 seconds (0:00:00), despite the user spending 45 seconds on the landing page:
And in the below example, the session duration is just 45 seconds, not 90 seconds. Because the user didn’t trigger any other hits on that 2nd page. No events. No other page views. No other types of GA data… so GA can’t timestamp that user leaving:
So here’s what how we treat this traffic…
As we’ve highlighted above, it’s not a reliable metric. As Google Analytics doesn’t provide genuine session durations. The average session duration metric for a given period actually includes all of these 0:00:00 sessions! This is why this metric shouldn’t be used as an aggregate KPI. As mentioned in the previous post, conversions & conversion rate beat session duration.
Look at landing pages by source (add a secondary dimension) – and analyse bounce & conversion rates next to those low session duration sources to truly get a feel for low performing sources.
If you spot a low performing source of traffic, with low session duration, high bounce and low conversion, then drill down further. Combine the source & landing page in a segment, and then analyse session durations, bounce & conversion by other dimensions. E.g. device category (desktop, mobile, tablet). E.g. by user type (new, returning). You might discover that new mobile users landing on a specific page are more likely to have lower session durations, and higher bounce, and lower conversions.
A landing page with lots of 0:00:00 sessions might have a video embedded on it.
The video might not be tracked/tagged for Google Analytics. I.e. might not be sending hit data when it is played. Which means that if a user lands on the page, and watches the video, but then leaves without doing anything else… then Google Analytics will record a bounce, and a session duration of 0:00:00. If that video is tagged, and sends an interactive event to Google Analytics when it is played, then you’ll get a more genuine session duration, and bounce rate, for those video-playing users in Google Analytics.
The same applies to an outbound link on the landing page. If the purpose of the www.example.com/landing-page is to get users to click on a CTA which takes them to a www.different-site.com to signup, then that outbound link should be tagged, to send an interactive event to GA. If not, you’ll never know if the page is doing its job properly.
