I'm planning on doing user testing and research to help improve the UI and UX of our portal/website.There are a lot tools to help improve your product. Think of tools like card sorting, user interviews, user tasks, etc.

I’m going to use card sorting, user interviews and some testing with the current version of our portal. A few of these tools can be used without the actual product. For example, you can do a card sorting test with users who have never seen the product before.


This testing will happen before the first release of our portal, so we do not have users. We're playing on asking potential users without prior knowledge of the portal. When I refer to 'users' I mean potential users we've asked to come in and test our portal.


Should you do user interviews and card sorting before or after user testing with the current version of our portal, like completing user tasks?

I can imagine that users might have more context when they know about the product but it might also cloud their opinion when they have knowledge of the product before the actual testing.

  • If you're card sorting then you'll need to have some content / info on those cards. Where are you getting that content from?
    – JonW
    Jun 30, 2017 at 9:00
  • We've been working on a first version of our portal for a few months now. All content on the cards is based on the version we have now and problems we've run into while doing some internal testing with colleagues. Jun 30, 2017 at 9:04

2 Answers 2


Hmm, interesting question.

Well, I was in a similar position sometimes ago when we were revamping the web portal. Your condition might me different but the time when we were catering that portal then that time it had almost 400 pages and the biggest problem was their navigation system including header, footer, breadcrumbs etc. and also the copies of the links.

We decided to do the reverse card sorting first and side by side we'll take the inputs of the concurrent users so that we can get the context and then while doing the card sorting we'll take help from both concurrent and the new users to get to a definite point.

Later we handled copies by doing AB testing.

The main point is we involved the manual presence of two types of users twice in making the system more prominent.

Your situation might be different but I hope this'll help.


First, it's important to understand what your users are like and how they perform the tasks that your app will help them with.

At my current company we do shadowing to simply observe users. We watch what software they use to do their work. We pay attention to the printouts on their walls. We notice how often they're interrupted and how frequently they're up from their desks. All that and more. From those observations it's pretty clear what needs to be built to support those workflows.

Observation provides immediate, unfiltered data, unlike interviews and surveys, which rely on self-reporting and recall.

Before you start designing screens you must figure out what to build, and these observation sessions help with that. Usability testing and card sorting are good after that.

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.