Eye Tracking at Juniper

Natasha Shimuk
Apr 25, 2017 · 8 min read

The Juniper UX team decided to explore how eye-tracking technology can enhance our user research. Below I share some of my first impressions from working with an eye tracker.

As a UX Researcher at Juniper, I try to view our products from the user’s perspective. When meeting with a current or prospective customer, I can learn a lot by watching how they interact with a product and then asking questions. Utilizing qualitative research methods like usability studies, interviews, field studies, and surveys always brings our team valuable feedback. But my sessions still fall short of seeing the product exactly as the user does. Using an eye tracker is one way to bridge that gap.

Our research team began to use one such tool in January. In this post, I’d like to introduce you to the exciting technology and then talk about a couple results from my first eye-tracking study.

How does eye tracking work?

It might surprise you to learn that eye tracking tools have actually been around for over 100 years. Today’s advanced technologies of course look nothing like the initial invasive and inexact experiments of the early 1900s (eye cups and levers and smoke drums oh my!).

Source: By Yarbus, A. L. [GFDL (http://www.gnu.org/copyleft/fdl.html) or CC BY-SA 3.0 (http://creativecommons.org/licenses/by-sa/3.0)], via Wikimedia Commons

The most common eye-tracking method uses pupil center corneal reflection. That’s a mouthful so let’s look at how it works on a very basic level.

When an eye tracker aims an infrared or near-infrared light at an eye’s pupil, the light won’t be detected by the eye, but it will reflect off of the cornea. The reflected light is then captured by the eye tracker’s camera. Based on the direction of the light vector that bounces off the cornea, the eye tracker can then calculate with great accuracy where the participant is looking.

Today’s eye trackers work in real time with video cameras and allow participants to sit and move their heads naturally. That’s important because as researchers, we want participants to feel at ease.

The eye tracker we used looked like this — a small bar that fits on the lower edge of the monitor.

Source: http://www.tobiipro.com/product-listing/tobii-pro-x3-120/

Before moving on, I want to introduce some additional terms that are important to understand when talking about eye tracking results:

  • Fixation: When a participant maintains their visual gaze on a single location, it is called Fixation. The duration of fixation is therefore the amount of time that a participant spends looking at that single location.
  • Saccade: When viewing stimuli, your eyes can move quickly between fixation points. That rapid eye movement is called Saccade.
  • Areas of Interest: When reviewing the results of a study, researchers can designate areas of interest such as images, text, and buttons within the broader range of stimuli on the screen. In the image below, for instance, each oval might represent an area of interest. Setting these areas then enables the tracking software to analyze the time to find and the duration of fixation for each one.
Source: http://www.tobiipro.com/learn-and-support/learn/steps-in-an-eye-tracking-study/data/digging-into-areas-of-interest/

There are two primary outputs from eye tracking studies:

Gaze plots are static images that visualize one or more participants’ gaze patterns through a series of circles and lines:

  • The circles indicate fixations and grow in size based on the fixation duration while the lines indicate saccades.
  • Numeric values in the circles indicate the sequence of fixations.
  • Different participants can be represented with different colors.
  • The primary function of the gaze plot is to reveal the time sequence of fixations: where, when, and for how long participants look.

Heat Maps show the distribution of fixation over the full range of stimuli.

  • The collected data are aggregated across all participants.
  • In the example below, the most frequently and longest viewed areas are indicated by red (followed in decreasing order by orange, yellow, and green).
Source: https://www.flickr.com/photos/rosenfeldmedia/2367261378

Why Should We Care About Eye Tracking?

To understand how eye tracking can assist UX research, let’s take a step back and think about what we’re trying to accomplish with product usability.

Good usability requires high marks for satisfaction, efficiency, and ease of learning for the user. Humans are creatures of habit: whether they’re opening a door or interacting with the interface of complex software, people expect and want their products to work a certain way. So designers can usually improve on usability metrics by incorporating the most common, shared habits.

Since an eye tracker can record what users see at a granular level not reached by human observation and interviews, it can help to identify and confirm shared habits. Here are a couple examples that you might already be familiar with:

No one likes a webpage filled with banner ads so it makes sense that people would try to ignore them as much as possible. But can we prove that that actually happens? Yes!

Thanks to eye tracking studies, researchers have repeatedly shown that internet users have learned to avoid not only banner ads, but also anything else that looks like one. In the below heat maps from Nielsen Norman Group, you can see that participants looked at just about everything on the pages except for banners.

Even valuable content displayed in the form of banners tends to get ignored because the human mind thinks it’s an ad.

Source: https://www.nngroup.com/articles/banner-blindness-old-and-new-findings/

Nielsen Norman Group also used eye tracking to illustrate one of their most famous usability findings: rather than read all text on a webpage, people would often read a couple lines near the top of the page and then scan down to the bottom. The eye tracker would generate a rough F-pattern (or something similar) in the resulting heat map. This behavior was most common on search engine results pages (SERPs).

Source: https://www.nngroup.com/articles/f-shaped-pattern-reading-web-content/

As Google (and Bing and others) updated their SERP interfaces, searchers’ habits adapted — and eye tracking again illustrated their behavior. After running an eye tracking study, Alex Birkett of ConversionXL wrote that, “With the advent of rich text and ad placement, users find themselves exploring the entire results page to find what they’re looking for. […] Many users look at rich text on the right before even considering actual search results.”

Eye Tracking at Juniper

In January our team began renting the Tobii Pro X3–120 eye tracker and got a license for Tobii Pro Studio, the accompanying software that lets us design and run studies, observe subjects live, save the recordings, and analyze and visualize the results.

At that time we were in the middle of redesigning Juniper’s online Network Design and Architecture Center, and we wanted to better understand how our users interacted with the information on our site. So I incorporated eye tracking into my usability study.

Here are two findings that the eye tracker helped us to better understand:

When I asked participants to find documents covering a certain topic, many struggled to complete a filtering task. We’d expected it to be fairly simple because the results automatically updated based on the filtering selections.

But there was a design flaw: the filter was above the fold while the results were below the fold! When participants didn’t see the updating results, they searched for a submit button that wasn’t there.

After the task, some of the participants articulated the problem while answering my questions. But the eye tracking results made the design flaw even clearer.

Take a look at the gaze plot below. There are large clusters of fixations in the empty top and bottom corners of the filter section. Participants expected to find a submit button in one of those places, and they weren’t sure what to do when it wasn’t there.

Additional eye tracking statistics showed that it took an average of 30 seconds for participants to scroll down the page and find the filter results, a section that I had designated as an area of interest. That is way too long!

Eye tracking also confirmed that we placed a page element in the right (expected) location. In a separate task of the same study, I asked participants to find the highest rated documents. Most of them quickly found the Sort By dropdown menu and selected Rating.

In the gaze plot below, I’ve circled the large cluster of fixations over the dropdown menu (an area of interest that I set prior to analyzing the results). In this case, it was particularly useful to see the sequence of fixations. While it’s difficult to see from the static plot with fixations piled on top of each other, the eye tracking software allows researchers to replay a participant’s experience as each fixation is plotted.

The eye tracking confirmed that the dropdown was not only quickly located, but that it was also one of the first elements focused on by participants.

Eye Tracking Enriches Our Understanding of User Behavior

Eye tracking can provide a piece of the usability puzzle we wouldn’t otherwise get; there are limitations to what I can observe and even what participants consciously understand about their own learned habits as they scan an interface or website.

But there are some caveats:

  1. Eye tracking certainly isn’t the be-all and end-all of UX research; talking with participants and observing their interactions with our products remain critical components of my research sessions.
  2. Like any new research methodology, there’s a steep learning curve related to the complexity of the collected data. Expect to invest additional time to understand how to run eye-tracking studies and interpret the results.
  3. We conduct most research projects remotely, but participants will need to come to our usability lab to participate in eye-tracking studies.
  4. Eye tracking won’t make sense for every study. For instance, if your tasks take a user through multiple pages in your site or product, it can get hard to connect the dots.

Ultimately, though, in the right situations, the more complete understanding I gain from eye tracking will let me make recommendations to my design team with greater confidence.

I’m looking forward to incorporating eye tracking into future studies.


When preparing to run my study, I found this guide from Nielsen Norman Group very useful. If you’d like to learn more about the eye tracker we used, click here to read about the Tobii Pro X3–120.

Juniper UX

User Experience and Design at Juniper - Blog

Thanks to Ridhima Gupta

Natasha Shimuk

Written by

UX Researcher @ Juniper Networks

Juniper UX

User Experience and Design at Juniper - Blog

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade