How we do usability testing – all in moderation

September 20, 2018

There are many ways to skin a cat, as the saying goes. There are also many ways to conduct usability testing (see ‘Why do usability testing’ to sample some of the different flavours).
For us, the best method is moderated usability testing in the lab or remote. Why? It allows you to find out why problems happen on websites and apps by watching behaviours in real time and probing around testers’ responses.

What is moderated usability testing?

Simply put, moderated usability testing requires users to undertake tasks using “think aloud” protocols under the supervision of a trained moderator, either in a lab or remotely using desktop sharing software.

Usability testing ingredients – the Web Usability way

When running a moderated session, we will have:

  • Testers who have been specifically recruited to fit the client’s required profile and so are real-life users – not ‘professional’ testers
  • Skilled moderators who adopt a user led approach to facilitation and are able to explore issues (often unexpected) as they arise, probe on why testers do things and their emotional reactions to the site
  • Eye tracking equipment so we can see what users do and don’t look at; whilst we don’t recommend heatmaps, ‘liveviewer’ adds enormously to the insight into how users look at websites and apps
  • Labs and observation studios so clients can observe the testing in real time – and ask additional questions ‘on the hoof’ prompted by the testing outcomes and not anticipated in advance

How to turn research into action

Knowing what the problems are is only half the battle. Identifying the right actions and getting them implemented is the difficult bit! Some criticisms of lab based usability testing are that it ‘involves numerous meetings and produces fat reports that nobody reads’. Well, not the way we do it.

To ensure this research leads to action, we combine our style of usability testing with a ‘sense-making’ workshop. Anyone who can influence the implementation of research outcomes observes at least one research session before developing a collective view of the issues to be addressed. Finally, we ask everyone to agree the appropriate actions in the light of the client’s resourcing, technological, and political constraints.

This helps us do away with bulky reports no one will ever read and ensures something actually comes of the usability testing!