When conducting user testing I sometimes find that participants describe how they will click on links within the prototype and talk about the actions they will do rather than actually physically clicking on the prototype itself and going through the prototype journey. I wonder if this is because there are no rollovers to indicate this is clickable? When a participant does not click on the prototype, what is the best way to encourage them to click and perform the actions rather than just describe they would click without creating bias?

  • Any app screenshots? Maybe the design is the issue.
    – Danielillo
    Commented Jul 27, 2022 at 5:59
  • I find this happens across different designs and different testing sessions. So I don’t feel it’s one particular design that is causing this response. Just what is best to do when this happens?
    – Sam
    Commented Jul 27, 2022 at 6:21
  • 1
    Maybe you need to improve the way you lay the basis for these sessions, be explicit that this is an interactive prototype they are looking at rather than a static mockup. Maybe they are coming from a previous experience with someone else that just used to present static mockups.
    – Adriano
    Commented Jul 27, 2022 at 6:39
  • @Sam Sometimes the answers emerge more easily and are more objective with a real visual example. Just sharing any examples you mention will help to better focus the issue.
    – Danielillo
    Commented Jul 27, 2022 at 6:40

1 Answer 1


I am assuming you are talking about moderated usability testing here. Setting expectations up front is your best bet - you can be as explicit as makes sense for the given test you are running.

In moderated sessions I tell people up front if what they are seeing is interactive, if it has certain limits, and I always encourage my Ps to think aloud as they work. If they just talk through something when I ask "How might you accomplish X...", I will sometimes just prompt them to go ahead and try what they just described and then ask if the results matched their expectations. I don't believe this approach has biased my results.

It might be a little trickier in unmoderated tests, but you just need to be clear in your instructions up front, and your prototype needs to be more "bulletproof": in addition to clear instructions and context, I will sometimes chunk out workflows into smaller steps, add optional remediation instructions at points along the way, and include an ever-present "Start over" button in case they get down a rabbit hole.

I will also sometimes add in large click zones over "non-interactable" areas of the screen with a message to the effect of "these controls are not part of this test", and take note of how often people try to interact with those areas, since it can tell you more about what they are expecting.

Of course this is user experience so the proper answer is always "it depends," but simple prompting, if done well and subtly, can work just fine. Actual mileage may vary.

  • 1
    That's a nice tip about the "non-interactable" parts of the screen. Thanks for that, I'll be using it. Commented Aug 1, 2022 at 6:48

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.