Eye tracking is the process of recording where a person looks, for how long, and in what order, using specialized cameras or sensors. It turns something people normally take for granted, a glance at a headline or a scroll past a button, into data a team can measure and compare.
Product, UX, and marketing teams use eye tracking to see past what someone says they noticed and look at what they actually noticed. That gap matters, because people often misremember or guess when asked to describe their own attention.
In this article, we’ll explore how eye tracking works, the types of tools available, what it can measure, and how to run a study that actually shapes design decisions.
What is eye tracking?
Eye tracking is a research method that measures the direction, duration, and pattern of a person’s gaze as they look at a screen, page, or physical space. An eye tracker records where the eyes point and how they move between points of interest.
Most modern eye trackers use near-infrared light and a high-definition camera. The light reflects off the cornea, and software calculates the exact position of the gaze several times per second. That capture rate is called the tracker’s frequency, and it determines how fine-grained the resulting data is.
The output is usually a heatmap, a gaze plot, or a video overlay showing exactly where attention landed. Researchers then combine that visual data with think-aloud commentary, click tracking, and body language to build a full picture of how someone engaged with the content in front of them.
How does eye tracking work?
An eye tracking session starts with a stimulus. That could be a webpage, an app screen, a product package, or a physical shelf display. The participant looks at the stimulus while the tracker’s sensors follow their eye movements in real time.
Two eye movements matter most here. A fixation is a brief pause where the eyes lock onto one spot, usually because something caught attention. A saccade is the rapid jump between fixations, and it happens too fast for the eye to register detail during the movement itself.
Software stitches these fixations and saccades together into a timeline. Many platforms also track pupil dilation, since a widening pupil can signal increased cognitive load or interest. The result is a dataset researchers can turn into visual reports, and increasingly, tools built for UX research fold basic gaze or attention prediction into the same dashboard as surveys and usability tasks.
None of this happens instantly on its own. Calibration comes first, where the participant looks at a few known points so the software can map the geometry of their eyes to the screen in front of them. Skipping or rushing calibration is one of the most common reasons a study produces messy, unreliable gaze data.
Types of eye tracking technology
Not every study needs the same hardware. The right type depends on budget, setting, and how precise the data needs to be.
| Type | How it works | Best for |
|---|---|---|
| Screen-based (remote) | Camera and infrared sensor sit near a monitor and track gaze without touching the participant | Lab-based usability testing, website and app studies |
| Mobile or wearable | Glasses or headset with built-in cameras track gaze as the person moves | Retail, in-store, and field research where participants move around |
| Webcam based | A standard laptop or phone camera estimates gaze direction using software alone | Large-scale, remote, low-cost studies where precision matters less |
Screen-based systems give the most consistent accuracy for digital product testing. Webcam-based tracking trades some precision for scale, since it can run with any participant who has a working camera and an internet connection.
Budget usually decides the starting point. A small in-house UX team often begins with webcam or predictive software, then moves to screen-based hardware once eye tracking becomes a regular part of the research process rather than a one-off test.
Eye tracking vs heatmaps vs mouse tracking: what’s the difference?
These three terms get used interchangeably, but they measure different things.
- Eye tracking records actual eye movement and gaze position, captured by a camera or sensor.
- Heatmaps are a visualization, often built from eye tracking data, that color-codes where attention is concentrated, typically red for high attention and blue for low.
- Mouse tracking records cursor movement and clicks, which correlates with attention on desktop but is a proxy, not a direct measure of where the eyes actually went.
In short, a heatmap is usually the output of eye tracking, while mouse tracking is a separate, cheaper method that approximates the same idea without hardware. Teams that cannot run true eye tracking often rely on heatmap and hotspot testing built from click and scroll data instead.
The Nielsen Norman Group’s original eye tracking research is a useful reminder of why the distinction matters. Their eye tracking study, not a mouse-tracking or click-based one, is what first identified the F-shaped pattern many people scan text in, a finding that later heatmap-based tools have built on rather than replaced.
What metrics does eye tracking measure?
Raw gaze coordinates are not useful on their own. Researchers translate them into a handful of standard metrics before drawing conclusions.
| Metric | What it tells you |
|---|---|
| Fixation count | How many times the eyes paused on an element, an early signal of visibility |
| Fixation duration | How long each pause lasted, often linked to interest or confusion |
| Dwell time | Total time spent looking at a defined area, such as a button or headline |
| Area of interest (AOI) | A defined region researchers compare against others, like a nav bar versus a hero image |
| Time to first fixation | How quickly an element was noticed after the page loaded |
A long fixation is not automatically positive. It can mean someone is engaged, or it can mean they are confused and re-reading the same line. Metrics need context from a task, a follow-up question, or a think-aloud comment to mean anything on their own.
Real-world eye tracking examples
Eye tracking shows up across several fields, each using the same core method for a different question.
- Market research
Brands use it to see which shelf position, package design, or ad element draws the eye first, before a shopper consciously decides anything.
- Website and product UX
Teams run eye tracking alongside a website usability checklist to confirm whether a call to action, menu, or form field actually gets seen.
- Healthcare and cognitive research
Clinicians study gaze patterns to support research into conditions such as Parkinson’s, autism, and dyslexia, since atypical eye movement can be an early signal.
- Human resources and training
Organizations study how experienced employees scan a dashboard or checklist, then use that pattern to train new hires faster.
Across all four, the point is the same. Eye tracking replaces a guess about what people look at with a recorded, repeatable answer.
A retail team, for example, might discover that shoppers fixate on a product’s price before its label, which changes how a package gets redesigned. A software team might learn that most users never look at a settings icon tucked in a corner, which explains a feature nobody uses despite good reviews for the product overall.
Pros and cons of eye tracking
Eye tracking is a strong method, but it is not the right fit for every research question.
Benefits:
- Captures subconscious attention that participants cannot accurately self-report
- Produces objective, quantifiable data instead of memory-based answers
- Works well alongside unmoderated usability testing, since participants behave naturally without someone watching over their shoulder
- Generates visual reports that are easy to share with non-research stakeholders
- Can run in a lab, in the field, or remotely depending on the hardware chosen
Limitations:
- Shows what someone looked at, not whether they liked it, so it works best paired with follow-up questions
- Screen-based and wearable hardware can be expensive and requires trained staff to run well
- Webcam-based tracking sacrifices some precision for lower cost and easier scale
- Small sample sizes can produce heatmaps that look conclusive but are not statistically reliable
How to run an eye tracking study: step by step
A useful eye tracking study follows a clear sequence, regardless of which hardware type is used.
- Define the question.
Decide what you actually need to know, such as whether users notice a new checkout button, before choosing a method.
- Choose the setup.
Fixed lab systems work well for controlled comparisons in a usability lab setting, mobile glasses suit field research, and webcam or predictive software fits fast, remote studies.
- Recruit representative participants.
A handful of the right users beats a large group of the wrong ones.
- Run the session.
Combine gaze data with think-aloud commentary and task completion, so the “where” connects to a “why.”
- Analyze the output.
Build heatmaps for attention concentration and gaze plots for the order in which elements were noticed.
- Turn findings into changes.
Pair the visual data with usability scores or survey feedback before recommending a redesign.
How QuestionPro supports eye tracking and usability research
QuestionPro does not sell dedicated eye tracking hardware, but it fills the gap most teams face after a gaze study ends: turning attention data into a full picture of the user experience. Heatmap and hotspot questions inside QuestionPro Research Suite let teams layer attention data with direct feedback, while QuestionPro UX tools support task-based usability testing for the same website or product being studied. Together, they help researchers connect what people looked at with what they actually thought about it.
What eye tracking really tells you
Eye tracking cannot explain why someone liked or disliked what they saw. What it does well is remove the guesswork about what actually got noticed in the first place.
That distinction is the whole value. A heatmap paired with a usability score or a short interview turns a simple attention record into a design decision a team can defend. Used that way, eye tracking becomes one input among several, not a replacement for asking users directly what they think.
Teams that treat it this way, as one piece of a larger research process rather than a standalone answer, tend to get design changes that hold up once they ship.
Frequently Asked Questions (FAQs)
Yes, but accuracy depends on the setup. Dedicated mobile eye-tracking glasses stay accurate as a person moves, while webcam-based tracking on a phone is less precise due to smaller screens and inconsistent lighting during a session.
Costs vary widely by method. Webcam-based or predictive software studies can run a few hundred dollars, while professional lab sessions with dedicated hardware and a trained facilitator often start in the low thousands per project.
No. Eye tracking shows where attention went, not why. Most research teams in the US pair gaze data with a short interview or survey question to explain the behavior the heatmap reveals.
No special training is required. Most sessions start with a brief calibration step, where the participant looks at a few points on the screen so the software can accurately map their gaze before the actual task begins.
No. Retailers use it for shelf and packaging design, researchers use it to study cognitive conditions, and HR teams use it to understand how employees scan dashboards or checklists during training.
