The purpose of usability testing is to test the usability and/or ease of use of a product or tool, and then make calculated changes and improvements guided by the test results.
You should always test your website with real and representative users.
The most popular methods of usability testing are
- user testing
- expert review
- A/B and multivariate testing
- first click test
User testing is a method by which users are asked to perform certain tasks in order to measure a website's ease-of-use, task completion time and the user's perception of the experience.
It is one of the best ways to find out what is or isn't working on a website.
There are 2 types of user testing: in person testing (face to face) and remote user testing.
- in person user testing: during a face to face usability test, moderators are seated near the test participants, observing the users behaviour through given tasks, answering their questions and observing what is happening on the screen and their non-verbal behaviour
- remote user testing: in this type of testing the participants carry out tasks from their home or place of work and the usability tester observes their behaviour using screen sharing technology
The following guidelines apply to both in person and remote testing.
- Plan the scope, issues, participants, location and budget
- what are you going to test?
- what concerns do you have about the site that you want to test?
- which users should participate in the test?
- where will you conduct the test?
- what is your budget for testing?
- Develop scenarios
- select relevant tasks for users to try
- prepare, try out and refine scenarios for those tasks. Make sure the scenarios are clearly written and that they can be easily completed in the allotted test time
- Recruit test participants
- recruit people that represent your current or potential user group
- consider using a firm that specialises in recruiting for usability tests
- if you do it yourself, build a database of users for future tests
- Conduct the testing
- make sure participants understand that it is the website which is being tested, not them – they are helping to show strengths and weaknesses of the current design
- get participants to describe their mental process aloud as they work
- let participants express their reactions
- listen, do not lead: be careful not to ask questions that may influence the participants' responses
- take notes on actual behaviour rather than inferences
- Document and make good use of the test results
- compile data from all participants
- list any problems that participants had
- sort the problems by priority and frequency
- develop solutions
- test the revised version to ensure you made the right design decisions
An expert review helps pinpoint usability problems in the design and interface of websites and software. It is used early in the design process, sometimes on paper mock-ups or prototypes, to ensure that design problems are not built in. Expert reviews can also be carried out on existing websites and pages.
The expert doing the review gives a rating according to how well the website or prototype meets general usability guidelines, including readability, accessibility, information architecture and content quality.
- Preparing the review
- identify the organisational goals of the website or prototype
- define key tasks that users are trying to complete based on input from the website/section manager, analytics, social media, trends and keywords and input from surveys and feedback forms
- decide on the principles on which the review will be based
- choose 3-5 reviewers, who are briefed on the goals of the website and the main tasks beforehand
- Conducting the expert review
- each reviewer works through the guideline checklist and records the problem, the page on which it occurs, and the relevant usability principles
- reviewers can also choose to talk through their review with the leading usability expert taking notes
- the process should not take more than an hour for each reviewer
Please find the supporting document Usability Expert review, which will help you to evaluate your website in an efficient and effective way.
A/B testing and multivariate testing
A/B testing, also called split testing, is a form of user test comparing the strengths and weaknesses of 2 different design versions of a single page. A/B testing is a powerful tool but is limited in that only one change can be tested at a time.
Multivariate testing, on the other hand, allows you to test multiple variations at once.
- choose exactly what you want to test
- develop a hypothetical situation
- decide how you are going to measure the success
- test group and split
- use randomisation
- document your methods and results
Please find the supporting document AB testing, which will help you to evaluate your design variations in an efficient and effective way.
First click test
First click testing looks at the first thing that website visitors click on, on a given page, in order to complete their intended task.
It’s a useful tool to assess the effectiveness of the linking structure of your website, including navigation. It lets you see how users get around the website and complete their intended task.
- define and document the tasks
- track each click the user makes
- time how long it takes the user to make this click
- after each task, use a scale to evaluate if users were happy with the information they found
- assess the ease or difficulty of completing each task
- document your test and results
Please find the supporting document First click test, which will help you to examine what a test participant would click on first on the interface in order to complete their intended task.
Any personal data that is processed as part of any task relating to the development of the European Commission's web presence must be done so in compliance with Regulation (EU) 2018/1725 of the European Parliament and of the Council of 23 October 2018.
See data protection for more info.
Contact and support
Need further assistance on this topic? Please contact the team in charge of Europa Domain Management (EU Login required).