Most passers-by never notice the display

What it is

A lit public screen is, for most people walking past, as if it were not installed. The durable finding is display blindness: the eyes can graze the surface while attention never lands. The problem is not that the content is dull. Passers-by file public screens with ads, signs, and lightboxes—objects it is adaptive to filter—while walking already spends vision on the path, obstacles, and companions. This entry is about whether attention hits the glass at all. It is not about why a noticer fails to interact, and not about whether anyone believes the screen can be touched.

Why it happens

Walking is a visual job. Optic flow, collision avoidance, and wayfinding occupy resources near the fovea; a vertical surface beside the path arrives in the periphery. Many deployments sit parallel to the flow, so heading never points at the screen. Geometrically it is not a thing being looked at.

A second layer is learned filtering. Glowing rectangles in cities have long been advertising. People acquire a skip-schema much like banner blindness. The more a screen looks like digital signage, the more that schema claims it. “Bigger will get noticed” therefore fails: size improves resolvability, not the classification “this is not for me.” Phones, shopfronts, and oncoming pedestrians compete for the same budget; a public display has not purchased anyone’s gaze.

Studying it

Count in the field; do not walk participants to the screen and ask whether they saw it. Code each passer-by from the flow: head turn, gaze beyond a glance, slowing. Intercept people who have just passed and ask what was on the screen, splitting seeing from encoding. Huang’s observations of public displays and Müller and colleagues’ work on display blindness sit here: not-looking is the majority act; a glance is the typical look.

Independent variables: mounting orientation (oncoming vs flanking), distance from the main path, density of visual competitors, whether the surface reads as an ad slot. Dependent variables: notice rate among passers-by, dwell of gaze, recall at intercept.

Asking people in a lab to “look at that screen” deletes display blindness. Outdoor eye tracking is rarely wearable in a crowd, so coding leans on visible head turns and stops and will miss peripheral registration. Notice rates inflate in the first days after install; take a baseline after habituation.

Where it stops holding

In concourses, lifts, and queues people are captive; the attention budget is higher than in a corridor, and blindness eases—without licensing the claim that corridor screens “are actually watched.” Information people are already hunting (departures, emergencies) will send eyes up; blindness applies to screens the current task did not nominate. A novel form (floor projection, a one-off sculpture) can briefly break the schema; once the class is common, filtering returns. A flanking screen in a fast corridor has almost no geometric chance of a look. That is placement, not a content failure.

Applying it

• Mount anything that must be seen on the heading of the flow, or at a node where people must turn (a corner, a gate, a wait point). Do not expect a wall parallel to walking to capture gaze.
• Design primary content for a glance: very large type, very few items. Do not use “they will walk closer” to rescue a screen that was never seen.
• Count, among people who passed, how many turned their head toward the display before talking about copy or controls. When notice is near zero, buttons will not be seen either.
• Verify in a stable flow window by coding at least several hundred passers-by for head turn and stop; intercept at the exit with one fact that should have been on screen. If both turn rate and accuracy are low, move and reorient before rewriting.

Related

• Within the group: K7.01.2 Noticing-to-interaction is a multi-stage leak · K7.01.3 Static content almost never yields interaction
• Adjacent: K7.02 Discoverability of Interactivity · Z8.01 Ambient displays and peripheral information · A5.16 Habituation and cue failure · A5.07 Attention capture
• Search terms: display blindness · public display · noticing · passer-by