Go Back

Comparing Funnel performance Mobile vs Desktop

July 3, 2018

Discovering what type of devices are responsible for converting is very important. What we are thinking about our target group and their behavior can be very different from what we can and should measure and check by analyzing web analytics data.

Recently the significance of mobile devices has increased in a big way. More and more often the visits are performed on a smartphone or a tablet, even in business-related topics. It seems obvious that website should be compatible with any type of device. How can we check if Users are able to perform the desired action without any disturbs though? We can compare the Funnel in terms of device used.

First we have to create necessary Segments which we are going to use during the funnel comparison. Let’s create segments which will filter all the data to Mobile and Desktop visits.

Mobile devices segment configuration:

Desktop segment configuration:

Having those two segments configured we can start comparing the funnels. To do that we should go to Comparison > Compare Segments in the main menu:

After that we have to choose desired segments for each funnel – here we are using configured segments. Fist funnel in the comparison will contain Desktop visits, second – Mobile and third – All visits to show the bigger picture.

When segments are applied, we can decide what we want to compare. In this case, we will go for Contact form Funnel which contains Contact page pagieviews and Contact conversion itself:

And here is what we can get with all this configuration:

Now we can easily compare how many visitors:

  • entered the funnel,
  • from where did they come,
  • where did they go,
  • how many visitors proceeded to the conversion,
  • conversion rate.
  • Thanks to this analysis we are able to detect if somewhere within the funnel Users are dropping too often using a particular device. If for example, mobile users are constantly leaving before filling the form it can mean that the web form is not showing correctly or the validation is wrong. In that case, we should immediately test this step on as many devices as possible.