Bargain basement usability testing

Paul Boag

There is a lot written about usability testing both online and off. However, except for “Don’t Make Me Think” by Steve Krug they all seem to make it sound very complicated and expensive. So here is the boagworld guide to bargain basement usability testing.

The boss refuses to pay for usability testing, you have no budget of your own and yet you are determined that the site will be as easy to use as possible. What do you do?

There is a perception that usability testing has to be time consuming and expensive but it really doesn’t need to be that way. The key is to do a little and often therefore catching problems before they become too complex to change.

Start off by doing some testing on initial design concepts or even rough sketches. Test again when you have a more detailed information architecture or a basic wireframe. Test yet again once the majority of content is in place and finally do one round before you go live.

That sounds like a lot of testing doesn’t it? Well it would be if the testing involved large numbers of participants who exactly matched the target audience and if they all were tested in expensive usability suites with cameras and observation areas.

However with bargain basement usability testing you make up for quality with quantity. By doing a lot of rounds you will probably pick up on more problems than just a single round done with all the expensive techniques.

For bargain basement usability testing all you need is 5 or 6 friends (or random strangers), a desk, a chair, a computer and a notepad.

Picking your users

Lets start by looking at the users who you will test with. First of all they don’t need to meet the exact demographics of your user group. It can be very time consuming to find people that match your audience. Instead just grab anybody. It doesn’t matter who. Friends, family, random strangers. The reality is that the majority of usability problems are going to be generic across all audiences. Sure if you have a friend that meets the criteria all the better but don’t get hung up on it. Better to test somebody than nobody at all!

Also don’t bother getting anymore than 5 or 6 people. More than that and you really don’t catch that many more problems. It just isn’t worth the additional time testing. You are better off just testing with a few, fixing the problems and testing again. That will reveal new problems which users never reached in the first round.

Your testing environment

Next lets look at the environment. There is no need to test in a usability lab. In fact I would argue the best place to test is in the location people normally surf the net. If possible go to their home or office and test on their own PC. That way they feel relaxed and are more likely to behave normally. If you cant do that then just sit them down in front of your PC and test them there.

Finally you don’t need a camera or observation area. Simply sit with them and jot things down as you go along. If you can get a colleague to join you and take the notes all the better but that is optional.

How to test

So what do you do in a test session? Well that doesn’t need to be rocket science either.

What you can do is largely dictated by what you have to show. When all you have is a couple of design concepts the best you can do is ask the user what they would click on to complete a specific task. You can also ask them if they can quickly spot certain screen elements using the same flash testing we have spoken about before.

As you have more to show you can ask users to complete more complex tasks. You can ask them to find a specific page within your wireframe or ask them what they would expect to find within a section on the information architecture.

Your job is to watch how they interact with your site and note where they struggle. The key to this process is to encourage them to articulate their thinking as they go along. Users will often sit staring blankly at the screen if they are stuck. You need to ask them what options they are considering and why they are struggling.

Before they click on anything ask them why they made the choice they did and what other options they rejected. The idea is to get them to talk as much as possible about what they are doing.

Fundamentally that is it. Obviously I could write a lot more on the subject but that should be enough to get you started. Get 5 or 6 random people to sit in front of a computer, look at your site and ask them to complete a series of tasks while talking about them. Not difficult is it. Why then are usability experts paid so much? ;)