• What is Eye Tracking?

Eye Tracking

Eye Tracking is the practice and technology of recording where on a screen, page, or scene a person is looking - and for how long. In digital design and marketing, eye-tracking studies generate heatmaps and gaze paths that show what visitors actually focus on versus what designers assumed they would.

Once expensive specialist research (lab-only, $50K+ studies). Now accessible via webcam-based services and AI-predicted heatmaps that approximate the same insights at much lower cost. Both have value when used honestly.

What eye tracking actually reveals

Three patterns it consistently surfaces:

The F-pattern of reading. On long text-heavy pages, eyes typically scan in an F shape - across the top, partway across the middle, then down the left side. Means content above the fold and on the left edge gets disproportionate attention.

Banner blindness. Visitors actively avoid looking at areas that match standard ad placements (rectangular shapes in known ad positions). This held even when the “ad” wasn’t an ad - important content placed in those zones gets ignored.

Face direction effects. Photos of faces draw gaze. The direction the face is looking pulls subsequent gaze in the same direction. A photo where the model looks at the CTA button increases CTA attention measurably more than the same photo with the model looking at the camera.

Where eye tracking is misused

Three patterns:

Treating heatmaps as recommendations. A heatmap shows where attention went, not whether the design was good. People may have looked at the wrong thing because the right thing was hidden. Heatmaps are diagnostic, not prescriptive.

Webcam eye tracking presented as lab-grade. Webcam-based services have improved enormously but still produce noisier data than dedicated lab equipment. Reasonable for directional insights; not reliable for high-precision claims.

AI-predicted heatmaps treated as actual data. Tools that “predict” eye-tracking results from a screenshot are pattern-matching against past studies. Useful for early-stage design feedback. Not the same thing as observing actual user behaviour.

What eye tracking is genuinely useful for

Two real use cases:

Pre-launch validation of critical pages. Homepage redesigns, key landing pages, signup flows. Catching attention misallocation before launch is much cheaper than catching it via conversion-rate decline post-launch.

Resolving design disagreements with data. Two designers disagree about whether the headline or the hero image is doing the work. Eye tracking provides ground truth. Cuts subjective debates short.

An example

A B2B SaaS team had spent four months redesigning their homepage. The new design tested at 1.8% conversion against the old version’s 2.3% - a regression they couldn’t explain because the new design “felt” cleaner.

An eye-tracking study (webcam-based, 32 participants) revealed the issue: the new design had moved the primary CTA from the top-right (where it had lived for two years and where users automatically looked) to the centre of the hero. The centre placement looked better aesthetically but conflicted with the F-pattern attention path. Users looked past the new CTA on first scan and often left without finding it.

They moved the CTA back to the top-right while keeping other improvements. Conversion lifted to 2.7% - better than either the old or the originally-redesigned version. The four-month redesign was salvaged by a 90-minute eye-tracking finding.

The aesthetic improvement was real. The aesthetic improvement at the cost of attention pattern wasn’t worth it.

Related terms