As the webmaster of SupplementReviews.com, I’m constantly analyzing data and trying to improve some key metrics. One of the most important metrics is the average session duration: how long the average visitor spends on the site.
No matter what I do, I’m always frustrated when logging into Google Analytics. Here’s why:
(Monthly Average Session Duration in minutes, per Google Analytics)
I’ve made substantial efforts to help keep visitors on the site, but no matter what I do, the trend just keeps going downward. Back in 2010, the average session duration was nearly 5 minutes, but now we’re down to just under a minute and a half.
Is it my site? Or is it just the changing nature of internet users themselves?
Number of visitors using a Mobile device is simultaneously increasing.
(Monthly device traffic share)
Now, the quick and easy assumption would be that visitors on mobile devices are simply spending less time on the site. And while this is technically true, we’re seeing a decline of session duration across all types of devices:
(Monthly Average Session Duration in Minutes by Device)
So there is more to the story than just an increase in mobile usage itself. Let’s dive into a few theories that might explain the data:
In order for a theory to make sense, it needs to fit two criteria:
A gradual change that,
explains a decreasing average session duration.
There’s a lot of theories one might have that involve changes to the site itself, but none of them would explain the gradual nature of the decline.
Design changes are usually deployed all at once.
From early 2011 to early 2014, I hardly made any major changes to the site at all. During this time, I was working with various contractors to rebuild the site from the ground up, so the existing design wasn’t being actively improved. Despite this, the average session duration was gradually decreasing.
Even after deploying a complete rebuild in April 2014, subsequent design changes were deployed in batches, not gradually.
The server was only upgraded 3 times.
One possible explanation is that the server got faster over the years, however the actual upgrade schedule doesn’t fit the graph at all.
We changed hosts 3 times in the last 7 years. Twice in 2010 and once in 2013. Again, this doesn’t fit the “gradual” requirement to explain the decrease in average session duration.
The content has only improved.
Since 2010, not only have we collected thousands more reviews, the editorial standards for these reviews has skyrocketed. Instead of single-sentence reviews, we’re now requiring 400+ word arguments made for or against each product.
Both the quality and quantity of our content has gradually increased, so it doesn’t seem logical to blame content for the change in average session duration.
Internet speeds have steadily increased.
There is no question that this has been happening gradually over the last 7 years. Faster internet means that visitors will get to the information they are seeking more quickly, thus complete their session in less time.
However, this would only explain a few seconds of change, not a 3-fold decrease.
The attention span of the average visitor is decreasing.
The term “attention span” might not be entirely correct. The more accurate explanation may be that visitors are more impatient, less willing to read large piece of text, better at scanning pages to find what they want.
While there’s no way to ultimately prove it, the theory that more and more visitors are seeking “instant gratification” seems to be a very likely explanation of the data we see.