Public Eye: Using Eye Tracking Software To Improve the Usability of Your Website
Jinfo Blog

By Julie Howell

Item

The web design industry is understandably excited about eye-tracking technology. Finally, technology that knows what the people who use your website are thinking!

Should you believe the hype? Eye-tracking software isn't cheap. If your company decides to invest in eye tracking, are you confident you will see a quick return on your investment? And does eye tracking provide all the testing you need to build your website's information architecture? Will eye tracking mean an end to other forms of user testing?

These are the very questions that we at web design agency Fortune Cookie asked ourselves during 2006. And of course, the questions cannot be answered with a definitive yes or no. But if used in the right way, eye tracking can definitely deliver results.

Here's the idea behind eye tracking: When we read a website -- or anything else on screen -- our eyes dart about, skimming and scanning in the hunt for killer content. We process information very quickly, acting on impulse, ruthlessly sizing up and dismissing content options.

Eye-tracking software records these eye movements, giving you information about eye motions that can be impossible to otherwise retrieve. Sometimes users don't remember where they look. And in interviews, users may not always tell the whole truth about what they looked at. Eye tracking provides a candid, unmediated account recorded in real time.

But like any usability technique, an eye-tracking study is only as good as the people who carry it out. It's only really useful if you know what you want to use it for, such as improving the flow of information on a webpage. And its findings are only as good as your ability to interpret them.

We at Fortune Cookie are committed to execute and interpret eye tracking research in the most effective way. Like everyone else in the web design and usability industry, we had read about the benefits of eye tracking on sites such as Jakob Nielsen's popular useit.com <http://www.useit.com/eyetracking/>. We knew that eye-tracking technology would impress our clients. But would it also deliver ROI? Is eye-tracking software anything more than an expensive gimmick? If users tend to read websites in the 'F-shaped pattern' that Nielsen describes in his articles, did we really need to purchase our own equipment to prove the point to our clients?

We think so. When used intelligently, eye tracking is a powerful tool to support usability testing and help maximise return-on-investment (ROI).

But before deciding if an eye tracking study is for you, start by finding out what eye tracking is and when it is likely to be most effective.

Why use eye tracking?

Eye tracking is not in itself a new methodology, and it can be tempting to want to use it on everything and everyone at every stage of the design-and-build process. But our experience shows that for eye tracking to deliver optimal results, and have a positive impact on ROI, it needs to be deployed wisely.

We use the Tobii Eye Tracker <http://www.tobii.com/> with ClearView analysis software. Eye-tracking software is used to chart the progress of a user's gaze across a computer screen. Prior to testing a website, the eye-tracking software must calibrate each individual user's eyes to ensure that the eye tracking works effectively. Although this may sound complicated and intrusive it is in fact a very simple process that takes a matter of seconds and is neither stressful nor difficult for the user. Once the calibration process is complete, eye tracking can be used to:

• Observe how the user's eyes wander across the design as he or she attempts to find a piece of information, complete a task or navigate through the site to a deeper level.
  
  
• Produce fascinating 'gaze plots' and 'hot spot maps'. These are graphical representations of the user's sight behaviour that show where the user looked on the screen, in what order they looked at items on the screen or at parts of the design, and for how long they 'gazed' at something prior to moving on. Such charts are extremely interesting as usually the user is completely unaware of how much their eyes look around a webpage before they decide what to read or which link to click. Web designers and site owners alike are usually amazed by the results. See Image.
  
  
• Provide documented evidence that can be used to inform changes in the design. This is extremely important if it is necessary to persuade whoever is paying for the site re-design to make additional funds available for the work. Charts or indeed video that clearly illustrate how much time a user spends trying frantically to find the information they are looking for are incredibly compelling. Web users are notoriously unwilling to spend more than a few seconds on a webpage trying to find what they want before giving up and trying a different site.
  

Time really is money on the web. Site owners, when confronted with eye-tracking evidence, quickly realise the implications of failing to make the user's progression through the website as swift and easy as possible.

• Provide information about user behaviour that occurs on a 'subconscious level' (such as eye movements). Traditional user testing -- that involves observing a person who expresses what they are thinking as they navigate a site -- cannot provide subconscious- level data.
  

At Fortune Cookie, we make a distinction between 'web designers', who are concerned with coding websites and producing the graphical content, and 'information architects', who work with clients and users to arrive at the most effective structure for the website. This latter group also look at the site as a whole as well as the 'webpage estate' (that is the positioning of links, text and graphics; use of 'white space'; the order in which content is displayed and how it is grouped, etc.).

Eye tracking is of greatest benefit to information architects (IAs) at several points in the design lifecycle, including:

• Evaluating the existing site
  
  
• Testing wireframes
  
  
• User testing new designs
  

Our IAs cite several advantages to using eye tracking over more traditional user testing. These are:

• It makes it easier to track process flows. In traditional user testing, mouse clicks can be observed and the user can be encouraged to voice what they are thinking as they progress through the site. Eye tracking, however, records user behaviour that neither user nor observer may be aware of. Behaviour that adds precious seconds onto the user's journey is behaviour that cannot easily be explained. Analysis of eye-tracking gaze plots and heat spot maps quickly reveal why the user has taken too long to find the information they were seeking.
  
  
• It is easier to spot the weak points in the 'funnel'. A website's information architecture should guide the user to the information they are seeking with minimum effort. This is the 'funnel'. If users are getting stuck or taking too long to progress through the site, the funnel must be altered and improved. Eye tracking can inform these improvements.
  
  
• It is a useful tool for comparative design. Particularly when working with a new client, it is useful and enlightening to show the client how their site performs when compared with a competitor. Comparing eye-tracking data between two (or more) websites can provide a strong indication of which sites are performing the best. When wireframe designs are produced, eye tracking can also be a useful way of choosing the design that is going to be the most effective.
  
  
• It can be used to improve the performance of particular elements of the site. One of Fortune Cookie's biggest clients is a financial services provider that wanted to improve the performance of some promotions on their website homepage. Fortune Cookie used eye tracking to measure the amount of time users' eyes rested on each promotion. We discovered that promotions that made use of photographs had a big impact, and the position of the promotion also affected the impact on the user. Traditional usability testing may not have revealed this.
  
  

User profiles

It is vital to recruit an appropriate number of the 'right type' of user for any kind of user testing. The agency supplying your users should produce 'user profiles' to ensure they're providing the most closely matched users to your client's needs (i.e. their target audience). A profile may describe the gender, age and any disability of the user. If an intranet is being tested then it is important to test with users who have different 'job profiles'. There is a law of diminishing returns when user testing, and as few as eight of the right type of participants is usually adequate to solicit useful feedback.

Structured questions

Preparing structured questions in advance of testing makes it possible to decide if eye tracking is the best way to solicit data or measure behaviour. Gaze plots (which show the sequence of glances an eye makes) are fascinating, but they are meaningless unless they reveal user behaviour that can influence changes to the site design. Therefore it is vital that IAs decide what it is they are hoping to measure prior to commencing the tests.

The questions need not be complicated, although there should always be a strong strategic focus to any eye-tracking programme. In the case of our financial serice provider client our structured questions considered the order in which users looked at items on the page and how long their gaze rested on the priority items.

Have a clear sense of the question(s) you want to address. Which of these designs works best? Does this promotion effectively attract attention? Is the copy clear and intuitive?

It is extremely unlikely that an eye-tracking study will be the only tool for testing the efficacy of a website. It should form part of a usability-testing toolkit that includes a range of tests, such as psychological questioning. Playing back a session and asking users why they looked where they did -- and why they didn't look where they didn't! -- can deliver invaluable insights. You may find an eye tracking study isn't even appropriate for a group of users.

Eye tracking: the bottom line

Correctly used, eye tracking can help deliver tangible ROI by helping developers come up with new designs that accurately reflect user expectations and behaviour. Your company will find that by keeping an eye on where your users look, you can expect increased sales and greater product awareness.

Related FreePint links:

• "Working with a Usability Specialist" By John S. Rhodes <http://www.freepint.com/issues/240806.htm#tips>
  
  
• "Books About Usability" By Dr. Jakob Nielsen <http://www.freepint.com/issues/100806.htm#bookshelf>
  
  
• "Analysing User Behaviour: A Case Study" By Chris Kutler and Ray Devaney <http://www.freepint.com/issues/080905.htm#tips>
  
  

« Blog