Multi-tab browsing and its impact on web analytics

Who doesn't remember having about 15 browser windows opened at the same time on their computer? I certainly do, and what a pain that was! Opera first introduced the tabs on its browser and unwillingly opened a brand new way of browsing the Internet.

I'm sure that I am not the only one opening 10 articles in different tabs when on a news website. A simple click from my mouse wheel and there it is, queuing to be read. I sometimes overuse it to the point that I end up barely reading half the pages I opened.

Most of the main browsers now include this option: Opera, Firefox, Chrome and even the latest version of Internet Explorer.

In web analytics, the visitor's path is a good way of knowing how people land on your page and where they go next. As you can see on this screenshot below, we know that most of the people coming on that page land directly from either direct traffic, either from a search engine or either from an external link pointing to this page (more commonly called referring site).

We also notice that most of the traffic leaves that page as well after reading the article and therefore shows I am having a hard time retaining that traffic.

I certainly am not an expert in web analytics but when I was having a closer and curious look to the visitors' path statistics on my own blog, I asked myself this question:
How can the path can be measured if people open different articles in numerous tabs?

If you have 10 unique visitors coming to your website, 4 of which leave directly after viewing the homepage and 6 of them browsing to articles, that gives you respectively 40 and 60% of your traffic as shown on the graph below. This results in 16 page views (10 for the homepage and 6 for the articles).

This is quite simple and do not require advanced skills to understand. However, my question may be better explained with the following situation. In the graph below, we have 10 unique visitors landing on the homepage, just like in the first situation. Four of them leave the website before browsing any further. Out of those 6 visitors left, let's say several open different articles in different tabs. For obvious reasons, the analytics tracking tag on your page will still count those tabs as page views. Out of the 10 unique visitors, let's say 3 of them clicked on the first article. This means that 30% of your visitors have viewed this article. If you add up those percentages, they go well over 100%, which used to be impossible.

This clearly shows that having the multiple tabs tool on most browsers benefits to your website as it encourages people to browse more and view more pages.

However, in terms of analytics, this quickly becomes an issue. Instead of having a visitor going from page 1 to page 2 to page 3, we now have visitors going from page 1 to page 2,3,4,5 at the same time just like on the graph below.

While multiple tab browsing improved greatly our user experience, I'd be interested to know how this has impacted analytics online and the visitor's behavior online in particular. I've searched for some information on the Internet but couldn't find anything on this topic apart from an article from 2006 so if any of you are familiar to this or have answers to enlighten me, please post a comment below!

Translate page

Rate this article

Liked my post? Share it!

Bookmark and Share

3 great comments!:

Bernardo said...

As I can see the tab only allows a more dynamic site navigation, and that can impact on the path and on the exit report. But only if a visit is ended with more than one tab open.

Andrew said...

I actually opened this post in a new tab about 5 hrs ago and I'm just getting to read it now! I wonder if having an inactive tab running in the background has any effect on GA's Average Time on Site metric?

GA's FAQ says "..However, Time on site can be misleading because visitors often leave browser windows open when they are not actually viewing or using your site."

They don't mention tabs though.

Hadrien Brassens said...

Thanks guys for the comments!

@Andrew: Concerning the average time, I am not sure if it's affected. In fact, I have noticed in Google analytics that there is a high proportion of visits that end up spending 00h00m00s on my site. Even if people bounce back, they should be spending several seconds. Unless GA does not count visits under a minute and the seconds are only shown as an indication for averages...

GA definitely needs to give us more detailed information about these things I suppose :)

Post a Comment