Simple Ways to Improve your Usability Research

Usability testing and user testing in general can provide valuable insights into your designs. Recently I’ve been working on some usability testing on some designs for a SaaS product. Here are some useful tricks I found to be helpful.

Plan Ahead

Planning any sort of user testing session requires time, however it is well worth doing as you’ll yield the most reliable results, and in most cases “winging it” just won’t do.

Planning begins by preparing the test mock ups or site; from this you need to create a test script and schedule for yourself. The easiest way to create a plan is to list what the aim of the user testing session is; this way you can focus your questions and tasks.

After this is done take some time to sanity check your plan and user test with someone who you won’t be testing on. This way you can amend the test as you see fit.

This planning part of the user test is really for yourself. The test participants are more likely then not to behave in ways you haven’t anticipated; this isn’t something to be disheartened by.

Think of the Logistics

When testing without a dedicated testing room, it is worth trying to conduct the user testing sessions in the same room, this way it will reduce the amount of time spent setting up.

Try to create the closest possible experience to the ‘real thing’ you can. You can do this a few ways:

  • Use a mouse (not a Magic Mouse unless the user actually uses them).
  • Chose a non-natural scrolling setting as this is typical a PC setting.
  • Create a task as close to real life situation as possible. If necessary add a time limit.

Recruiting your Testers

There are many ways for recruiting users for usability testing. You’ll need someone with at least basic knowledge of the service you are designing for.

In my experience I found this to be a little difficult when dealing niche or complex software services. The easiest way to find people is to test internally; a bonus to this is that you don’t need to deal with any compliance or security issues.

When it comes to how many users you should test on I found that the Nielsen Norman Group’s recommendation of 5 has worked best. By only testing on five users per test you’ll spot the biggest usability issues in your design and allow for a quicker interactive design process.


  • Create a script for yourself, this will create consistency throughout your tests.
  • Sanity test your plan and testing session with someone you won’t be testing on.
  • Try and test in the same room — It’ll be easier for you.
  • Test the equipment before hand — You don’t want tech failing you when you should be testing.
  • Mimic the conditions and scenarios the user might face.
  • If it is difficult getting external testers, use your companies employees.
  • You only need to test on 5 people per test.

About the Author

Marwa Khalil is a London based UX and UI designer who has had the chance to work with some great London startup including Nutmeg and Peak. She is a regular contributor and editor for the LTUX London blog. Stay up to date with her on Medium and Twitter.

Originally published at on August 6, 2016.