User testing

Research and testing

Back to services

Analysis and testing are often supplemented with end-user testing. The aim of user testing is to find out how useful and intuitive is the site, and what kind of attitudes the users show towards it. During the test, potential users are performing typical tasks. We closely observe what kind of difficulties the participants are facing, what are the typical mistakes they make, what are the tasks that require effort and what are their opinions on certain solutions. When they finished the test we ask them about their expectations. Based on the test results we propose suggestions how to modify and improve the site to best serve the client’s goals.

Website usability tests are not carried out in focus groups. We do not ask people about their opinion of the website, because users think of various things which don’t necessary coincide with the problems they experience during the interaction. User testing also differs from professional testing where trained testers examine the site. They focus on programming bugs, and have a different perspective from everyday users and they can cope with things that would mean serious obstacles for average users. If something works it does not necessary mean that it is easily usable.

Steps of test:

  • Identify the weaknesses of the site
  • Defining the most important features for the user
  • Preparing the test
  • Choosing participants
  • Conducting the test
  • Evaluating and analyzing the results, discussions

The test should not last longer than one hour because the process requires continuous concentration from the participant. During this time period the subjects have to complete three or four tasks. Ideally, user tests should be carried out with minimum six and up to ten participants to highlight the most severe problems.

During the test, the subject is working on an everyday browser (with one he or she is familiar with), while the test moderator takes notes indicating the errors and the problems that occur. Task completion time should also be measured, as well as the level of completeness. We can also record the screen and the user actions (for example: Camtasia, Silverback, etc.) and the user’s remarks as well. We should encourage the participant at the start of the session to think loudly and verbalize everything he or she does or think. Afterwards, we analyze the notes and recordings, and summarize and weight the errors. We suggest solutions for severe problems. Don’t forget that an error that occurred only once can be also relevant since we are working with a small sample. The problem can affect even 10-20% of the users.

How to identify errors during the test session:

  • The user is unable to complete the task, gets stuck or try to solve it in the wrong way
  • The user gets frustrated by a feature
  • The user cannot find the solution for a task
  • The user acts slower than expected
  • The user misinterprets or is unable to interpret the feedback and communication of the system, or understands them with great difficulty
  • Information from client:Presentation of known problems, and existing research results
  • Client cooperation:Provide participants (optional)
  • Deliverables:Summary in Excel sheet regarding the problems and the potential solutions