clicks vs visits: a neverending story

by Luca Rabboni


Today I’d like to write about a question that should be pretty basic for those working in the  Digital Analytics world, but that unfortunately is often still too unclear: why is there a discrepancy between the number of clicks and the number of visits to the site?

Before starting to list the reasons for this difference I think is good to remember that clicks and visits are two different metrics and that they are calculated differently and above all they are collected by different tools (respectively adservers and analytics tools).

Obviously the assumption for this post is that your analytics tool is set correctly and that the discrepancy between clicks and visits is not a technical error.

Once I’ve done these premises we can start with a nice list of all the reasons :) :

  1. A sessions in Google Analytics lasts number of interactions plus 30 minutes of inactivity, so if I in this period I click several times on a banner, GA record only one visit against a total amount of clicks. For example, if I click on a banner and I land on a certain website then I click the back button and  again on the banner because I forgot to check something on the site I totalized 2 clicks and 1 view (at moment we have already totalized a discrepancy of 100%!).
  1. If I click on a banner and then I close the landing page before it executes Google Analytics script, it will be not recorded any visits despite I collected a click.
  1. Users may have chosen not to be tracked; in this case not every click will match a visit.
  1. Redirects or Pop-up may affect the proper tracking of a visit.
  1. The request for accepting cookies treatments can block the tracking of a visit until the user accept it.
  1. If a user has landed on your site after being exposed to a campaign and then returns within 6 months from a source that is not paid, this will be assign to the campaign source. In this case a click generates two visits.
  1. Generally Adservers (or Adword) are set in order to eliminate the clicks that are not considered reliable (e.g. from bots). GA instead records the visit.

In conclusion we can say that there are several reasons why you can detect a discrepancy between clicks and visits, but, beyond  techniques motivations, the message that I would like to go with this article is that this are two totally different metrics tracked by two different instruments. There are no benchmarks for which a 15-20% variation is acceptable, while a gap of 60% indicates some problem.

Obviously if there is a big difference you should go in-depth, but the only way to do that is to use your data as a benchmark and analyze any spikes or anomalies; certainly not trusting to some given data