Usability tools and techniques
A full version of this article is available at http://www.managers.org.uk/ (written, July 2003)
Usability tools and techniques are based on common sense and can be easily applied to your website. Lots of them can be performed ‘in-house’ and cheaply. Knowing the rules of usability tools and techniques before deciding to break them will enable you to approximate which types of customers may no longer visit your site. Consultants can charge you a fortune for something you could have done yourself. Be wary. For the most effective use they should be applied regularly, even built
into your daily routine.
Methodologies
- Usability: the goals of designing an interface (or website) to be effective, efficient and satisfying from the user's point of view
- Accessibility: is the above as well as taking into consideration that users may 'read' web pages with other types of technology that do not handle graphics or plug-ins (screen readers, old browsers, WAP enabled telephones)
- User-centred design: the process of involving your users in the design of anything you are going to produce i.e. a website
- Usability engineering: the process of using specific usability guidelines (e.g. W3C) to be made as explicit requirements and then measuring their implementation
- Human-computer interaction: the discipline concerned with the design, evaluation and implementation of interactive computing systems for human use and with the study of major phenomena surrounding them
- Interaction design: the design of everyday products to support people in their lives
- Usability metrics: guidelines to measure your return on investment
- Web metrics: the process of taking data from your web log files and using it to benchmark how many visitors you are getting on your sites
- User and Stakeholder Needs Analysis: Consultants analyse user and stakeholders’ needs (people within your company) often by collecting information using questionnaires, interviews, and then they present the answers in the shape of a report. It can be expensive and time consuming.
- Techniques used for the above may include:
- Task analysis: where a user performs a specific task (e.g. find some information on your website) and the ‘expert’ analyses and records what they do in order to complete the task
- Icon intuitiveness testing: users are asked to identify what an icon means
- Card sorting: 3 x 5” cards with various proposed website page headings are placed in front of users who are asked to sort them into related piles so that a generic navigation system can be identified
Planning, feasibility, and requirements
Design
- ‘Cognitive Walkthroughs’ are when users talk aloud as they use your website and explain their actions. This information is captured by the ‘expert’ to enable them to build up a picture of how your website is navigated.
- Prototyping is the process of mocking up a proposed website and then showing it to people in order to get their opinion.
- Low fidelity prototyping is a paper mock up of the proposed website.
- Medium fidelity has the top level implemented and one or two levels down to test certain aspects the proposed site.
- ‘Wizard of Oz’ is a method of testing a system that does not exist, normally using some form of prototyping.
Evaluation
- Usability testing may take place in purpose built usability laboratories where the users are recorded and watched through a two-way mirror. Alternatively testing make take place ‘field-study’ or ‘ethnographic’ style, these terms just mean in typical working conditions i.e. your office.
- Usability testing is when a group of users or ‘experts’ test your site according to specific guidelines. These may be based on several sorts of guidelines.
- Scenarios: a user given a specific scenario e.g. go to this site and buy a downloadable e-book.
- Heuristic: experts have a list of guidelines (or heuristics) and they evaluate whether the site meets these guidelines. This type of testing can be biased and is sometimes called subjective evaluation.
- Usability reviews quite often are the same as heuristic evaluations.
- Usability inspection (again, same as above).
Competitive analysis and benchmarking.
Competitive benchmarking is the process of analysing competing websites and deriving some statistical information from the analysis and then using that information to try and recommend improvements to your website. The disadvantages of this approach is that getting hold of the competitors statistics can be difficult and the statistics that companies who offer these sorts of services can be inaccurate by a large margin.
Do-it-yourself usability tools and techniques
Any of the above techniques can be performed by you. You do not need to go to a specialist. Decide what it is you need to know about your website and then think practically how you can implement these tests and what you hope to achieve by them. Testing your site using such techniques should be something you do regularly in order to constantly strive to improve your websites.
Software
- There are many types of software that you can buy to test usability in a more sophisticated manner. These softwares can be expensive and overkill for a small website:
- Click capture: Software that records where your user clicked so you can build up a pictures of how users surf your website
- Visitor timing software: During user interaction with the pages, the selected events types are captured, time-stamped, and recorded in session-specific log files.
- Eye tracking software: Software to capture what your users look at when they scan the pages of your website
- http://validator.w3.org/ is a free validator which will check your HTML code
- http://jigsaw.w3.org/css-validator/ another free validator which will check your style sheets.
- http://zing.ncsl.nist.gov/WebTools/ has visualisation software you can download to see how users surf round your site and which pages they browse.
Simple Testing
Testing is a must for a good website and can be done remotely (e.g. email your friends and say “Can you look at my website and tell me what you think honestly?”). Or locally (e.g. borrow a video camera and get your tester to sit in front of the camera and click through the pages whilst you say absolutely nothing and see what they do). Recording it helps you play it back and think about the design and how obvious it is.
- Testing with one user distinct from yourself or company is much better than not testing at all. Only four people can be sufficient for testing and getting feedback about your website.
- Testing early versions of your website is much better than waiting until the end. In this way you can fix what is not right before building on it.
- It doesn’t matter who you test with. It can be your family, you friends, etc., just as long as you are getting some sort of feedback
- Testing is there not to see what is wrong or right it is a tool for informing your decision making and design process. When designing and developing sites you can no longer see a specific site objectively in order to make the right decisions about it.
- Testing is iterative.