4 ways to make (or break) a user test

May 30, 2017
 
by
 
Adi Murphy
,
 
UX Consultant
 
, in
 
Research
hand using a pen to write checklist

The user test is an essential step in the design process. However, if you don’t take a robust approach to research, you may waste your time and budget. Worse still, you may receive misleading results which take you in the wrong direction. After conducting hours of hours of user research we've learned a thing or two about what works and what doesn’t.

Choose your participants carefully

Whether the research is moderated or self-moderated, it doesn’t matter. Who you choose is critical for success. For a well-established brand, consider contacting their existing customers. For a new product/startup then sit down and discuss who their target audience is. Once you know who to target you can source them accordingly. Your research will be irrelevant if you choose people who'd never use the product outside of the test. (Learn more about the importance of choosing the right subjects.) If you conduct regular user testing then try to broaden your pool of testers. Better yet, recruit new user test participants every time. The going rate seems to be around £1 per minute or £40 per hour. If this proves difficult, there are companies dedicated to finding people for research. Keeping your participants fresh is key. Testers who participate regularly will often think they know what you're looking for and act accordingly. The whole purpose of studying user behaviour is to understand how they would behave, perceive and feel about your product.

A well-written discussion guide

When moderating a test you'll save yourself a lot of time by having a carefully structured discussion guide. In a self-moderated test it’s imperative you write a good brief. This and the questions you ask are your only way of guiding the user through the tests. Every good story begins by setting the scene - think of your test as the story and the brief as your background.

Clear and coherent questions

The only way to keep a self-moderated test on track is through the questions you ask. If they aren’t clear your participants are going to struggle and go off track. Potentially you’ll end up with no usable results. In a moderated test you’re able to provide guidance when necessary. But if your questions are clear to begin with, you'll save yourself a lot of time. Plus less intervention means more authentic results. Writing good questions for your tests takes practice but it’s a worthwhile cause in terms of results. (Learn more about writing test questions.)

A thorough analysis and clear presentation

Your research is only as good as your analysis of it. And your analysis is only as good as your presentation of it so get that user test right! And your analysis is only as good as your presentation of it. Your results are worthless if you don’t present them in a way that your client will understand. Be objective. Don’t focus on the negative - include the positives too and vice versa. You want to give as clear & balanced a view of the product as you can. Focusing on what you think the client wants to hear does them a disservice. They need to hear about the good, the bad and the ugly. And remember - if it all goes horribly wrong you can always hire us to do it for you!

Adi Murphy

UX Consultant

Adi has been at the heart of the Natural Interaction team since joining us in 2013. Adi’s speciality is empathising with users and ensuring that our design doesn’t lose its humanity. She’s participated in interaction design and usability testing for clients including Jaguar, Consortium and Rolls-Royce.

Get in touch

To find out more about how our UX services can help your business, contact Alex now.

Contact us