01Overview
Change blindness is the failure to detect even large, obvious changes in a visual scene when they occur during a brief interruption — a cut, a blink, a flicker, or a moment of distracted attention. The change is in plain sight. It simply doesn't register.
For interface designers, this has a blunt implication: shipping a redesign does not mean users will notice it. Moving a button, changing a colour, restructuring a nav — these are invisible to users who are focused on completing a task, not scanning for what has changed.
02Detailed explanation
Simons and Levin's 1998 "door study" is the most striking demonstration. An experimenter asked a pedestrian for directions. Mid-conversation, two men carrying a door walked between them. During the two seconds of obstruction, the original experimenter swapped with a different person — different height, different clothes, different voice. Roughly half of participants didn't notice. They simply continued giving directions to a stranger.
- The "one-shot change" paradigm: two near-identical images are shown in rapid alternation with a brief grey flash between them. Participants can look at these images for minutes without noticing changes that are instantly obvious once pointed out.
- Changes to objects in peripheral vision, changes to low-salience elements, and changes that occur during saccades (the rapid eye movements between fixations) are especially likely to be missed.
- In usability testing: users will often interact successfully with a redesigned interface without noticing that the design has changed — which is not a validation of the redesign, just an absence of disruption.
03Why it exists
The visual system does not record everything it sees. It samples the scene — capturing high-resolution detail at the focus of attention, lower resolution in the periphery, and almost nothing during rapid eye movements. Between samples, the brain relies on a working "sketch" of the scene rather than a faithful store of all pixel-level detail.
Attention is the bottleneck. Everything outside the current focus of attention is processed at low fidelity — which means changes outside that focus often go unregistered entirely.
04Effects on users
- A button that moves to a new position in a redesign may continue to be clicked in its old position for weeks — users have a motor memory of "where it was."
- A changed error message may be functionally invisible if the user's attention is on the form field they're correcting, not on the message area.
- Users completing a task with high cognitive load will miss more changes than users casually browsing — the higher the task demand, the narrower the attentional spotlight.
- Loading transitions between pages can mask changes: what "flashed" or "loaded" during the transition simply doesn't register.
05Effects on designers & teams
- Redesign confidence: product teams often conflate "no complaints about the redesign" with "users are happy with the redesign." Silence is not approval — it may simply be change blindness.
- A/B test contamination: if the test variant changes something users aren't attending to, the test is measuring the absence of disruption, not the quality of the change.
- Accessibility reviews: visual changes that seem large to a designer who built them may be nearly invisible to a user scanning for a specific element.
06Practical takeaways
- Don't infer success from absence of complaints: users may simply not have noticed the change. Ask them directly — "what looks different to you?" — in post-launch research.
- Announce meaningful changes explicitly: if you've moved or changed a core interaction, tell users. Don't rely on them to notice.
- Test with task-focused participants: usability test participants who are doing tasks will exhibit more change blindness than participants who are asked to "look around." Match your testing to real usage patterns.
- Use animation to signal change: when an element changes state (a field validates, a count updates, a step completes), animate the transition — it creates a visual event the attention system can detect.
- Avoid critical information in the periphery: status messages, error indicators, and progress updates placed in peripheral positions are most susceptible to being missed.
07Design examples
The silent rollout
Shipping a redesigned nav without a changelog or in-product announcement means most users will not notice the change — until they look for something that has moved and can't find it. Silence is not the same as acceptance.
Inline validation timing
Validation errors that appear in a fixed position away from the active field are often missed when users are focused on typing. Change blindness is minimised when the feedback appears adjacent to — or as part of — the element that changed.
State transitions
A cart count that simply updates from "2" to "3" may not register. A count that briefly scales up, flashes, or animates signals change through motion — which the peripheral visual system detects even when the focus is elsewhere.
The missed change
If a participant completes a task on a redesigned interface without commenting on the new layout, this tells you the change didn't disrupt them — not that they saw it. Follow up: "did anything look different to you compared to last time?"
08Ethical risks
Change blindness is exploited when interfaces make consequential changes that users are unlikely to notice — updating terms of service without surfacing the change, quietly reducing a feature's capabilities, or moving a cancellation button to a less prominent location during an update. The change is there; users simply won't find it.
If a change matters to users, they deserve to know about it. Change blindness is not a loophole — it's a design responsibility.
10Suggested reading
Suggested reading is temporarily unavailable. Please check back later.