When doing user tests I discovered a recurring mouse path among testers. When asking them during a task to go back to the previous screen, the mouse was 'zig zagging'. The mouse went to the main menu first. After that to the close button and finally to the back button. See image below.

In an ideal situation, the mouse path would move straight to its target. That's clearly not the case now. I could be looking too deep into this. Testers were new to the system and hadn't seen it before. I had trouble finding research about mouse paths and what it means.

Is this something to look into more? Can we actively try and improve mouse path flow in our UI?

enter image description here

  • 3
    It sounds to me like user are looking for a link to the previous screen in the menu, then thinking that they might be able to get 'back' by closing the current screen before finally noticing the back button. Jul 5, 2017 at 9:22
  • @AndrewMartin Yes, true. My question is if this is something to try and improve upon or is it something that happens because testers are not familiar with the product yet? Jul 5, 2017 at 9:23
  • 1
    Maybe it's their way to think (I mean they buy time this way to consider the page) and realize where the back button is or should be, since they don't actually click on the close button for example... Jul 5, 2017 at 9:42
  • 2
    As @Jack-in-the-box says, I think it's just them discovering the location of the back button, so nothing to be concerned about. If the same users are doing it after continued use, then you might have a UX issue :)
    – user101673
    Jul 5, 2017 at 10:08

2 Answers 2


What you've discovered here are a few areas of interest for your users. Inferring an attempted use from those pieces is rough, but doable. To me, it looks like your users are familiarizing themselves with your navigation as the first step to entering the page.

Mouse tracking and eye tracking are heavily correlated (some studies say they're identical to each other around 88% of the time). In this way, they are both useful in defining areas of interest to the user. You can produce many of the same outputs from mouse tracking as you can eye tracking-- heat maps, for instance, are a popular product of mouse tracking.

What's more, mouse tracking has a few benefits over eye tracking:

  • It's cheaper, as there is no extra equipment to purchase or set up.
  • It's less susceptible to the Hawthorne effect, as mouse tracking can be done anonymously.
  • It's easier to discover missed interactions. A mouse can click; eyes cannot.


  • Regarding the benefits: a) there are solutions without extra equipment - just via webcam b) Hawthorne effect - absolutely right c) I don't get this point. With eye tracking you'll see the point where a user look and don't click. This is what you want to see. Not where interactions happens. To identify problems you should look onto the information which leads to interaction - or doesn't. And for eyes the pendant for the click would be the fixation - e.g. is it read? Both methods are valid and provide great insights. But on different levels and different effort.
    – JonnyZoo
    Jul 27, 2017 at 11:29
  • a.) A webcam is extra equipment, isn't it? You can capture mouse tracking with nothing but the computer itself, regardless of webcam availability. B.) Agreed. C.) I mean to say that you can see where a user may hover momentarily without clicking or interacting with a piece. This is useful for discovering what pieces users may have hesitation completing, or where they expect tool tips. You cannot gather this information from eye tracking alone.
    – denveruxer
    Aug 1, 2017 at 17:11

I don't have much to say and would comment, but I'm unable due to reputation:

I could imagine this pattern to occur in your UI because the back button is not located in the top left corner (as usual). The users might spontaneously navigate there before noticing the true location of the back button.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.