Digital

5 tips for usability testing from Steve Krug

June 18, 2015 by Jono Ellis No Comments | Category Digital Public Services, mygov.scot

This is a post by Kate Saunderson, one of our user researchers.

Is your website usable? That is the question that Steve Krug, author of ‘Don’t make me think’ and ‘Rocket surgery made easy’ continually asks. He wants others to ask as well, just as much as he does. So how should we answer that question and what tips does he have for carrying out usability testing?

Well, in a great piece of luck Steve was actually able to tell us himself, as he attended a recent session hosted by Government Digital Service. In short, it’s because user testing is often overlooked. Steve talked about approaches and actually went as far undertaking a test right there and then in front of everyone. Looking at the NHS Choices website, he recruited from the audience, looking at the scenario where someone will be looking to find advice online when their child has chicken pox.

So, what were Steve Krug’s 5 tips for usability testing? 

If your website is not usable by the intended audience, why continue develop it? The best way to understand if it is usable is by asking people to try and use it, observe what happens and identify problems to address. So, here are his 5 tips for usability testing.

Tip 1 – Define your driver for doing usability testing. If it is asking if people like your website, stop there. If it is asking if you have built something that people can use, continue. A well designed usability test should not seek feedback or opinion, rather it should try to seek insight into how usable your participant finds the site and identify problems for you to tackle.

Tip 2 – Decide who your audience is. Depending on the purpose of your site, you may have multiple audiences. So work out what you are looking at, and ask yourself who is most affected, start there. Steve highlighted that in the beginning you shouldn’t let your fear of recruiting properly stop you from recruiting at all, as this is a common area where people struggle and give up. It is better to do testing with somebody than nobody.

Tip 3 – Keep numbers small, don’t be ambitious. Three users per round is best, you can then choose to do as many rounds as you wish. As an observer, it is a manageable amount to observe, as a facilitator you can maintain focus. When people question whether three people is a valid representation, reply confidently, that your aim is not to prove anything, rather identify problems that can be fixed.

Tip 4 – Acknowledge the constraints and highlight the benefits. Asking somebody to come into a lab, undertake tasks and thinkaloud as they use your site is artificial, so is it valid? Yes it is, there are varying levels of distortions involved in asking a participant to leave their natural environment, engage with tasks, talk about how they are progressing and yet it still provides insights to problems. It works.

Tip 5 – Think about how you frame problems with the team. For small problems, it is best to talk about tweaking rather than redesign, you should focus on incremental change. Though serious problems can be resisted and big problems that go unaddressed can develop reputations. Sometimes you have to take the initiative to interview the team, understand why these problems were not addressed and use that understanding to lever change.

Sadly, there were no snacks for the audience, even though this is on his checklist for a good round of usability testing – a checklist that also tackles: create short sessions, get people to observe and debrief with them, use that data to create a short (2x A4) summary of tasks, significant problems and interesting things. If you want to go the next level, you can create video clips of identified problems.

We throughly enjoyed the session and participating, so a thank you to Government Digital Service, who kindly hosted the day.

We’ll be sharing updates on these features, and much more on social, so follow the team via @mygovscot on Twitter for more updates. Want to comment? Let us know below!

Tags: user research

Comments