Running usability testing on the cheap
Usability testing is important, and there’s no arguing with that. It’s a large part of design thinking and materialization to be specific. A step which can turn an OK idea into a great product and by omitting which you’re risking losing users, money, time or all of the above. So why aren’t all companies big and small running it? Most of the time the problem lies in the available resources, especially when it comes to start-ups. Some new people may get confused, so I feel that there’s a need to stress out that usability testing ≠ user research (sometimes called user testing).
User research involves interviewing the audience to establish whether they need your app and/or if they’re likely to pay for a pro membership along with other questions looking to identify any gaps between the business idea and
Usability testing is about finding flows in the user interface and fixing them.
This post is a simplified guide to usability testing on the cheap, in which you can use your family members, friends, or even co-workers, as long as they’re not a part of the development team!
Why is usability testing important?
Let’s start with the basics, what’s so damn important about usability testing? A lot of major UX design problems can be corrected by running even most simple usability testing, talking numbers, 50% of problems will be eliminated after the second test. Of course you’re going to have some slight UX troubles left over but it is very unlikely that major issues new users tend to encounter will survive both tests.
When is it best to conduct usability tests?
Usability tests are best conducted twice, once between the paper prototyping and a rough mockup stages, and once more during the stage of final development before shipping.
How many people will I need?
All you need is 3 people; a tester, a moderator and an archivist.
A tester is a user, whose job is to carry out a specific task using the interface, and to think out loud while doing so.
A moderator is in charge of giving users directions as to what they should accomplish, as well as communicating with the user and asking them leading questions that will prompt them to speak their mind out and help the archivist to track more problems. The users will not always be honest about having trouble searching for certain things or carrying out activities, so it’s the moderator’s responsibility to ask what’s wrong when the user seems to be having troubles by either reading the user’s facial expressions or looking at what they’re doing.
What’s moderator not to do is giving the user clues on how to carry out a task, it is important to be there only to encourage the user to give more feedback, not to teach them how to do it, else all will fail.
And last but not least, an archivist. Archivists are supposed to keep track of every problem the user is experiencing without ever getting involved into the process.
What’s moderator not to do is to be frustrated with the user, or trying to get involved into usability testing in any way. The archivist should always be only seen behind the scenes.
The perfect set-up
Ideally, the moderator and user are in the room, and the archivist is behind a glass or a wall or another form of separation. If the archivist is in the room they must make no noise or interact with the user or moderator in any way.
It is usually important to also place a camera that will be recording the user’s journey because some things will inevitably slip from the archivist’s attention.
Usability testing can be carried out by one single user
If you cannot afford the lab, or even 2 people out of your team, this can actually be carried out by one single user. In that case the user has to be proactive and speak out everything they’re struggling with and experiencing on a camera, that will be recording how they’re interacting with the app/user interface. The screen recording might work too depending on the software. You have to see what the user is hovering over and how much time everything is taking.
How many times does the testing need to be done?
When it comes to numbers, of course the more users you test the more potential UI problems you will uncover, but generally speaking 3–5 users should be enough for each testing. There will usually be at least several common problems all of them struggle with which will help you eliminate the biggest problems in the usability of your interface.
