UX Magazine

Defining and Informing the Complex Field of User Experience (UX)
Article No. 1183 February 5, 2014

Combining Contextual Inquiry with Eye Tracking

Traditional methods of user research and requirements gathering have served UX practitioners well in shaping our designs and redesigns of products and services. But as systems become more complex and users go beyond the screen, it’s becoming increasingly difficult to get accurate data on why our users, customers, and employees are doing the things they do.

Contextual inquiry has been a researcher’s best friend for a while now—a tried and true research method that helps collect insights from the field without having to rely on the user’s memory or ability to articulate their emotions and actions. However, we are now finding that what users say they do and what they actually do not draw a complete picture of what is happening.

Contextual inquiry comic

Because the design of many systems influences the eye movements of its users, using contextual inquiry along with eye tracking technology can surface answers to many deep-rooted and difficult questions. We can deliver gaze plots and look at heat maps, but more importantly, we can research what busy people are doing without interrupting their daily chores or connecting any recording gear to their highly secure IT systems.

Tobii X2-30

Tobii X2-30 monitor set up options

Historically, eye trackers proved to be obtrusive and clunky and required a long setup time. The user was required to sit in front of a special monitor or wear a headgear. This impeded the user’s natural behaviour and meant that the gear could not be inserted into someone’s workplace. Tobii’s new generation eye trackers are very small and unobtrusive and require minimal setup. This means that users can go about their normal activity without any interference from the equipment.

Why Look Through Your Users Eyes?

In a recent call center study, we used eye tracking to help visualize call center behavior and combine quantitative findings with contextual inquiry. It enabled us to uncover subconscious eye movements of the call center operators while they were busy using the system. Peculiar patterns of eye movements indicated issues with the system at several different levels: speed of data entry, navigation, and layout. Call centers are high-speed, patience-sensitive environments where operators are under constant pressure to get it right the first time for customers. Any issues with the interface can significantly slow down their process, affecting customer experience.

We were working with a complex, high-speed, transactional system. Operators needed to execute rapid micro-interactions, some of which they may not even realize they are performing. This particular system required a couple of weeks training to gain a beginner’s level of skills. The stressful environment of hard-to-use software and frustrated customers on the end of the phone resulted in high staff turnover, which translated to high costs for replacement, training, and productivity, straining levels of customer satisfaction for the organization.

With gaze replays provided by the eye tracking software we could see that there were numerous instances where the user’s eyes did not follow a natural and efficient flow, leading to severe strain on their eyes and reducing the overall efficiency of their ability to perform their jobs. Not to mention, it stressed them out!

They needed to scan all around the screen, AND refer to crib sheets to support them in order to find the required info in a timely manner. Imagine if you were a delivery driver and instead of planning your route at the beginning of your day based on the delivery locations, you were assigned jobs in random locations. You would be wasting more time and fuel than if your route was optimised based on bundled areas for pickup. The same thing was happening with the eyes of the operators, instead of 6-8 hours on the phone the business changed maximum shift times to only 4 hours because their staff was getting exhausted.

The findings from contextual inquiry with eye tracking became the basis for the design decisions

The gaze plot visualisation below was generated from the eye tracker. It shows the eye movements of two operators (green and pink) within the first nine seconds during separate customer calls. The dots represent the points where the eyes fixated and the lines represent the path between fixations (saccades). The information on the bottom left corner of the screen was not considered a part of the most common workflows, yet it was uncovered that the operators still looked at it at the beginning of almost every call to identify key system status information.

Design Best Practices Become Performance Worst Practices

An eye tracker placed under a call center screen showing gaze paths for two different operators

The Results

The operators and the stakeholders were surprised when they saw the results. Without eye tracking, it would have been much more difficult to establish that the problems were with the interface and not their operators. The results resolved some internal disagreements and budgetary questions. The IT team that managed the call center had been skeptical about watching operators at work, but eye tracking allowed us to understand their experiences without distracting them from the task at hand.

When the techies saw the eye tracker gadget and the outputs showing them the operator’s experience on live calls, it immediately evoked empathy. The decision-makers realized there was a need for change. The users’ eye movements we captured in the contextual inquiry research were used to directly dictate how the system was re-designed; both the information architecture and the hardware (keyboard).

How We Used Eye Tracking During Contextual Inquiry

Eye tracking can tell us where the user is looking but not why. Therefore, it is important to pair eye tracking with qualitative insights from the user as well.

To collect deeper insights, we used the Retrospective Think Aloud (RTA) method. With RTA, we showed the user the replay of their activity with and overlay of their eye tracking gaze. We then asked them to describe what they were doing while watching the replay. This helped trigger their memory of both their conscious and subconscious behaviors and also helped them notice fallacies in the system that they might not have been aware of while performing the activity over and over again everyday.

Design Best Practices Become Performance Worst Practices

Using a Tobii eye tracker during contextual inquiry

When the operators were shown how they were constantly looking to the bottom-left corner of the screen, they finally revealed that they were afraid of missing out on the constantly changing time-sensitive information that was displayed there and of facing the wrath of customers if they performed the wrong transaction. We noticed that some of the customers were not forgiving at all and screamed profanities at the operators.

The call center operators had automated many of their workarounds for the system and would have otherwise had a tough time trying to describe what they actually do. Without eye tracking, even a researcher observing the operator would find it difficult to tell where the problems were hiding because everything happens so quickly on the screen and the operators had automated their thinking, it was unconscious.

Conclusion

During this contextual inquiry, we identified pain points in the user’s workflow and the worked collaboratively with the users to provide realistic and compelling suggestions for improvement to process efficiency and effectiveness. For example, identifying what operators looked at the most before, during and after a call and placing those elements on the screen prominently made the operators’ eye gazes much shorter and they reported much less stress.

The findings from contextual inquiry with eye tracking became the basis for the design decisions that were made afterwards. More importantly, the eye tracking results were used to validate assumptions and get stakeholder buy-in. A lot of design teams spend hours arguing about where the users look. Eye tracking data helps end all those arguments and move forward towards what matters most, a better experience for the users.

ABOUT THE AUTHOR(S)

User Profile

James Breeze has a Masters of Organizational Psychology and his goal is to improve people's lives through improved design and usability of things. He runs Objective Asia and Eye Tracking consultancy in Singapore and SE Asia. Objective Asia was set up in February 2013 and is a subsidiary of Objective Digital, a UX consultancy in Sydney, Australia.

An eye tracking evangelist, he is also a Tobii Eye Tracker partner in Asia Pacific.

Add new comment

Comments

Hey James! Fantastic article!
It is something that we often tell clients, but then rarely do, because of several complications.
So usually we end up giving them a colorful scenario to do something and why they are doing it, but at the end we still work in a very simplified reality.

I think your methodology is very interesting for this "industrial" area, where you are testing this kind of software. How would you do it for "normal" websites or personal software like MS Outlook for example?

Cheers,

Rob

Hey Rob,

If you don't have access to a lab (or the funds), check out Userlytics. They have robust monthly packages for user testing on any device. But their user testing is not just video screen capturing while the user thinks aloud with each task, but it also records the user (in their normal environment) during the test!

Basically, you get eye tracking results with think aloud feedback, facial expressions and body language, without the large data sets. Instead, you get a snapshot report with double the footage.

In my professional opinion, this method kinda trumps eye tracking for usability purposes, however, only in the sense that you can test users on the go, i.e. mobile testing, and see their every response to your system.

@James

Great article! Well written research like this, coupled with ROI, is exactly what the world of UX needs!

Best wishes,

Leah

Cheers Rob!
For 'normal' sites or personal software we would probably do it in a lab with the same X2-30 eye tracker on a regular monitor. We could go to people's houses, and have before, but that is costly. I'd hope that MS has already done eye tracking usability testing on Outlook :)
See ya!