5 mistakes you should avoid when conducting usability tests

By 28th July 2014 Testing

Face it. Any app, website or various widget your team produces will eventually be judged by the harshest critics of them all. The public. One step before that comes a series of usability tests, which help you to prepare and fix issues in time for public release.

The unsuspecting masses on whom you’ll unleash your software decide whether all those sprints you put in to complete the product were worthwhile or not. Whether you actually crossed the finish line (i.e. you reached your product goals) or you missed the starting gun entirely.

It’s a risk that every creator has to contend with. But like any other risk, proper preparation prior to launch can successfully mitigate the uncertainty you face and increase the odds of having your product welcomed by its target users with open arms.

What are usability tests?

Usability testing is a key component of the development stage of a product. It is basically a rehearsal for the big day; you release unfinished copies of your product to a select group of users who get to try it and tear it to pieces, only for you to pick it up and make it even better than ever.

There is an art to conducting usability testing properly. Any good test designer knows that there are common mistakes that should be avoided to maximise the utility of usability testing and obtain crucial feedback that can make the product a runaway success.

In this article, you’ll learn about the five commonest mistakes which plague usability testers, and which you will be able to avoid when conducting your own tests.

Mistake #1 – Conducting tests late in the development process

There’s a tendency in the business to forget all about usability testing until after the last line of code is safely locked in place. Some put usability tests as a part of acceptance testing, They believe (wrongly!) that interrupting development work will increase the time spent on the task and hinder their flow.

I beg to differ.

Conducting usability tests at an advanced stage can actually increase the overall time spent on production. Testing nearly finished software is more likely give developers a harder time since any problems discovered will probably be lying deep inside layers of code that has to be picked apart and re-written all over again.

By testing usability from an early stage and in smaller chunks of unfinished software, any bugs that crop up can be swiftly dealt with, and the entire process suffers from less of a setback.

What to do instead:

Avoid bombarding your product with a barrage of usability tests at the end of the development process.

Start testing early and space out usability tests over the entire work process. A good idea would be to carry out usability tests at the end of every sprint.

Be careful to adapt the timing of the test to the schedules of the users as well. From our experience at ReQtest we found that tests typically take about an hour to perform, so it shouldn’t be too difficult to find an adequate timeslot for users to participate in testing.

Mistake #2 – Your user sample doesn’t match your target users

Just like showing up at a black tie party in swimming trunks is an embarrassing (though admittedly funny!) scenario, testing software with people who don’t accurately represent your target group of users can be the cause of some major chagrin come launch time.

It’s easy to grab your best buddies or team member and have them test your product every time, but chances are your friends are highly unlikely to make good ambassadors for the rest of the population. Especially if they all like coding or testing like you do.

How to avoid it:

To spare your blushes, we suggest that you give yourself plenty of time to find a good mix of people to make up your user sample. A good test needs adequate preparation time, so don’t rush it.

Naturally, you must ascertain key demographic data about your target users to be able to simplify and restrict your search to find only the people you want. This should be no needle in a haystack situation for you.

If you’ve dabbled in creating personas, make full good use of them as a basis for selecting and recruiting your ideal test subjects.

Vary the people you assign usability tasks to. If you are used to asking your friends, tell them to refer you to other friends and rope in new faces to test your software.

Mistake #3 – Poorly planned tests

By now I assume that all your tests are scheduled with reminders to give yourself ample time for preparation. You also have a group of users who are eager to test out your work and can’t wait to get their hands on it.

Is it time to relax yet?

No.

One question. How well have you planned out the actual usability test? Conducting a test requiresclearly-defined objectives that guide you to look for and capture all the important data you need to improve your product.

A common mistake is to carry out a usability test without deciding how it will help you achieve the specific purpose of enhancing your creation.

Tell me what to do.

Ask yourself a question. Several of them actually, such as:

  • Do I want to explore the details about the user’s experience of interacting with my product?
  • Am I looking to assess how well my app/website/software responds to the users’ demands?
  • Does my product fulfil its goals?

Each question requires a different test design and a corresponding data type to be collected.

You will know what kind of test to execute only after asking yourself about the purpose of the test in relation to the main project. Knowing what kind of test you’re planning to conduct also helps you assess the best type of environment where to carry it out and find a suitable location for the test.

Mistake #4 – Using bad testing methods

Nearly as bad as carrying out a test in an inappropriate location is conducting a usability test without the tools you need to collect the type of data you’re looking for.

Much like eating soup with a fork, a bad choice of testing software can starve your project from the feedback it needs to sustain the team efforts into making the end-product a better experience for the target users.

Okay, got it. What should I do then?

Choosing good testing software is of the essence for your project. You need to ensure that your data collection tool can give your reliable and accurate results.

The tool has to be simple and easy for your test subjects to use. Giving participants some time to familiarise themselves with the software is always a good idea

For the test itself it is preferable to use open questions rather than closed yes/no answers since the former yield richer data.

Also, resist the temptation of interrupting the participants often to tell them what to do. If that’s happening then the problem is either that the test questions are ambiguous or not well-written, or that the testing software is too difficult to use.

Obviously,as a test leader you ought to be very comfortable using the testing software yourself, as well as the data capturing method you adopt for the usability test.

Always prepare yourself sufficiently before going into a test. If you have the time, try conducting a smaller pilot test can help you discover any problems that have to be dealt with and fixed before the actual test occurs.

While testing isn’t exactly party-time for your participants, there’s no reason to approach the process as an anxiety-provoking, stress-filled event. Don’t let the stress of dealing with a situation you may not be familiar with get to you; after all, usability testing is extremely interesting because what you’re doing may very well impact the product in may positive ways.

Making people feel uneasy or upset is a sure-fire way of spoiling the usability of your usability tests.

While developers might chalk this mistake as a minor one, your attitude during a usability test can have a significant impact on the quality of the feedback you get.

I’ll simply put on a smile, right?

And try putting one on your test participants as well while you’re at it.

It’s important that you help users to feel relaxed during testing. This minimizes errors and increases comprehension of the instructions given.

You should explain the purpose of the test clearly and if necessary, write down the instructions so that the users can read them at their pace.

I think approaching a usability test with a good attitude is one of the most important factors in a test leader that contribute to the success of a test. And a truly positive attitude (as opposed to fake cheerfulness) is possible only when all the other points mentioned here are fulfilled.

Apart from a positive attitude, a common problem is that may be too stressed, resulting in you pointing out the solution instead of waiting for the test person to find the solution themselves “Just press the darn button, it is there, why don’t you see it”, is not a correct attitude to have as it undermines the testing process.

Conclusion

In this article I discussed the five commonest mistakes that you should avoid when conducting your usability tests.

To recap:

  1. Ensure that you give yourself plenty of time to plan your test and try timing it in the development process
  2. Have a comprehensive user recruitment schedule that helps you find the high-quality participants that give you high-quality feedback in your tests.
  3. Choose your tests wisely. Ask yourself the question your test seeks to answer and then design a test that can actually provide a definitive answer.
  4. Take care that the users are at ease during testing and that your recording instruments are working well.
  5. Finally, keep in mind that a positive attitude can have a big impact on the results of your usability tests.

Join the discussion One Comment

  • […] Another way to get usability measurable is to set standards. If there is a documented company standard, you can state that the system should be built according to the standard. Of course a standard stating that the OK button should be placed to the right of the Cancel button does not automatically mean that the system gets high usability, and the only way to really know if the system is ok is to perform usability tests. […]

Leave a Reply