If you contrast the old computer operating interface of the 1980s and 2010s, we have evolved from
to this,
Usability is about making system intuitive and simple such that users will automatically understand what it does. Today, we make icons or user interface (UI) elements similar to something that of real life elements. This is called as Canonical Representation.
International standards have also been established for Human Computer Interaction (HCI) and usability. The ISO 9241 Standard covers:
- The use of the product (effectiveness, efficiency and satisfaction in a particular context of use).
- The user interface and interaction.
- The process used to develop the product.
- The capability of an organisation to apply user centred design.
Why is user experience important?
For commercial platforms is boils down to conversions and sales. For intranets, usability is a matter of employee productivity.
- Every £1 invested in improving your website’s usability returns £10 to £100
(source: IBM) - A web usability redesign can increase the sales/conversion rate by 100%
(source: Jakob Nielson) - Usability methods raised user satisfaction ratings for a system by 40%
(source: Human Factors) - $1 trillion a year was spent on IT worldwide and up to 15% of IT projects are abandoned and at least 50% of a programmers’ time during the project is spent doing rework that is avoidable. (source: IEEE and ROI of User Experience)
When to work on web usability
The best time to do testing is throughout the design stage. This is because many of the problems you encounter would be structural in nature.
Correcting these problems after the development stage would require major changes to the existing architecture or database design. The longer you wait the harder it is to justify going back and redoing everything.
That being said, there is never a wrong time to test. So it you haven’t started, you should start planning to do so during your next project phase.
Where to conduct usability testing
You don’t need a dedicated usability laboratory unless you are doing very frequent testing. For most purposes, a quiet room is good enough.
Natural settings like a quiet room may be better as participants feel more comfortable and less inhibited.
The best practice we use internally is that we spend a good 15-30mins to warm up to the participant before immersing the user in the task.
How to conduct web usability testing
You should plan for quick and fast usability testing throughout your entire design phase.
As you progress through the course, you will learn many methods for conducting usability testing. The most basic and one of the most useful is qualitative user testing using paper prototypes.
To identify usability problems, having 5 users to test is typically sufficient. After each user testing is done, the design is rapidly enhanced in an iterative manner. This is iterative designing. We recommend at least 2-3 iterations.
You also may want to conduct a field study to see how your users are using your system in their natural habitat.
What to start qualitative user testing on?
- Your old Website – She is a good starting point to identify what are the good stuff that should be kept or emphasize and the problem areas that are giving your users pain.
- Your Competitor’s Website – She will provide a quick and dirty way to get excellent insight on the range of alternative interfaces and feature sets that you can adopt
How to do basic qualitative user testing
- Line up at least 5-6 representative participants for a 45 minute session, over a couple of days.
- What are representative participants? For example, if you are testing your new office intranet design, grab another colleague from a different department.
- If you are building a web application for customers 30 to 45, male and female, then get 2x males age roughly 30, 38 and 45 and 3x females with a similar age spread and relevant interest.
- For paid participants, $50/ session would be reasonable for a general user profile. Highly targeted users e.g. doctors may require $200-300/session.
- Factor in “no-show” by participants of typically 10%-15%. Preference is to call a day before hand to confirm.
- Allow 30 minutes in between sessions
- Place a computer in a quiet room and ask your participants to perform representative tasks
- What are representative tasks? For example, if you are testing your new office intranet design, frequently used task might be to search for contact information from a different department, book a meeting room or submit a form application.
- There are many user testing software that can help you record a video as the participants perform the tasks. This is helpful as it reliefs you of the need to take notes and should the situation arise that you have to persuade a client or your boss, this recording can be a useful tool by saying “hey look! It’s a random user’s opinion not mine.”
- Shut up and observe what the users do, where they succeed
- Identify recurring themes, for example if majority of your users encounter a problem, note it down.
- Prioritize the issues based on the severity it impacts your user experience. Clearly something that frustrates your user so much that he feels like returning immediate to Google search is an important red flag.
