Public ambient displays serve the passing many, not a single user

What it is

A public ambient display sits in shared space, but its "user" is not an operator seated in front of it — it is the flow of people passing by. The design object changes from "one person's session" to "foot traffic": content must be readable within seconds of passing, to an indefinite crowd, with no assumption that anyone will stop.

"Serving the passing many" fixes the content baseline: priority goes to what the crowd shares — next departure, current waiting time, exit directions — not to anything personalized. Putting personalized content on a public screen is not just a category error but a privacy exposure: individual-relevant information broadcast to everyone present.

It is a different medium from the phone screen. The phone assumes a willingly attending operator; the public ambient display assumes a crowd that has purchased no attention. Type size, information density, and interaction expectations all start from different premises.

Why it happens

Why public displays must be designed for passing:

• Traffic structure: movement through the space is the norm, stopping is the exception. The display's expected viewing time is the few seconds of passing, not minutes of standing.
• Attention budget: walking itself consumes visual and cognitive resources (navigation, obstacle avoidance); what is left for a roadside screen is a glance, not reading. The content budget of one screen is one or two fixation windows.
• Viewing geometry: distances in a crowd run from a few to tens of metres; type size, contrast and hierarchy are set by the farthest effective viewing distance — a parameter set with no counterpart on a handheld screen at arm's length.

One further environmental factor: a public screen competes for attention with the entire street — signage, crowds, advertising. Most passers-by develop display blindness, filtering public screens much as they filter ads: seen, but no attention allocated.

Studying it

• Attention to public displays: Müller and colleagues' field studies of public screens — observation plus intercept interviews on whether passers-by look, for how long, and under what conditions. Not looking is the majority behaviour; glancing is the dominant viewing form; display blindness is a stable finding.
• Interaction-stage models: Brignull and Rogers' work on large public screens distinguishes passing by, viewing, and interacting, and describes the honeypot effect — someone already interacting markedly raises the probability that others will. This gives behavioural measures for how a display moves from catching attention to drawing participation.
• Deployment studies: place a display in real space for weeks and observe the adaptation curve as novelty wears off; lab settings cannot produce this data.

Methodological caution: lab participants know they are "supposed to look at the screen"; real passers-by do not and should not. Gaze rates and content recall can only come from the field; there is no conversion factor from laboratory values.

Where it stops holding

• Waiting areas are not passing scenarios. Platforms and departure lounges hold people for minutes of captive time and can carry denser content and longer reads — corridor screens and waiting-area screens are two different designs, not one budget.
• The shared-need set is small: departures, time, directions, status. Beyond this set (advertising, news, community notices) viewing rates drop sharply — though their value may never have been in being read (atmosphere, sense of place); evaluation goals must be set separately.
• Glance rate is not the only metric: an ambient display's contribution to the feel of a space does not show up as "read the content". Evaluating atmosphere displays by content recall kills them wrongly.

Applying it

• Derive content from "readable in passing": first-tier information glanceable — large type, high contrast, three to five items per screen; details and second-tier content designed for standing readers, placed low or in an interaction layer.
• Set type size from the farthest viewing distance and verify on site; check contrast under worst-case light (direct outdoor sun), not showroom conditions.
• Align the display with the dominant flow's field of view (along the path's tangent), not perpendicular to traffic — side-facing screens get far less gaze than oncoming ones.
• How to check: fixed-point observation beside the flow, sampling glance rates; intercept people just past the display and ask one first-tier question ("how many minutes to the next departure?") — the correct-answer rate is the delivery rate. If delivery fails, cut information first, then check position and size.

Related

• Same group: Z8.01.2 Information must be understood in a brief glance, not a standing read · Z8.01.3 Ambient displays must not require interaction for basic information · Z8.01.4 Too many ambient displays become visual noise in public space
• Nearby: Z1.02 Peripheral and centre attention · Z8.02 Touchless services and public terminals
• Search terms: ambient display · public display · display blindness · glanceability