What dwell and desire-line data actually tells a curator

Curators already read rooms. They know which wall holds a sightline, which case needs more air, which label is too long for the light. What they usually lack is evidence at the scale of a whole visit: not one afternoon’s anecdote from a gallery attendant, but a quiet, aggregate picture of how the collection is really experienced — what people linger on, skip or return to, and which routes they actually choose through galleries and grounds.

That is what dwell and desire-line data is for. It is not a scoreboard of “engagement” in the social-media sense. It is behavioural evidence about attention and movement, mapped to objects, rooms and viewpoints, so curation can follow something better than a guess.

Dwell is not popularity

Raw dwell time is easy to misread. A long stay at a case can mean fascination, confusion, a bottleneck, a school group, or a bench that happens to face the wrong way. ARTT’s value is not a single number; it is pattern across many visits: clustered dwell that creates a pinch point; thin dwell in a room that was meant to be a destination; return visits to an object whose first interpretation clearly failed to land.

When Artty walks beside visitors — offering the way and the right story at the right object — those movement and attention signals stay on-device as personal experience and come back only as aggregates. A curator can then ask useful questions. Which objects hold a crowd? Which rooms get skipped? Which interpretations land? The answers support re-hanging, resequencing and rewriting labels with evidence rather than vibes.

A concrete pattern: a side room sees almost no dwell. The desire line bends around it. The entrance looked open on the plan but felt like a dead end on the floor. Re-hang the entrance — change the first thing the eye meets — and the room finally gets read. Another: one case draws long, clustered dwell that creates a pinch point. Give it more space and a second interpretation; ease flow without losing the moment that made people stop.

Desire lines are the hang visitors vote for

Architects talk about desire lines as the paths people wear into grass when the paved route is wrong. Galleries have them too. Visitors do not follow the sequence on the wall text; they follow sightlines, fatigue, the pull of a bright room, the push of a dark corridor, the rumour of a famous object two rooms ahead. Desire-line data makes that informal voting visible.

Mapped across a wing, desire lines show where the intended narrative breaks. Mapped across grounds, they show which outdoor routes are actually used. Mapped against content — which stories Artty offered and which were skipped — they show which interpretations land. That last link is what turns analytics into curation: not only where people walked, but whether the story at that spot held them.

The same language applies outside museums. On a waterfront, a desire line that bypasses a row of merchants is usually a fixable routing problem, not a verdict on the shops. In a downtown, a quarter two streets off the main route stays quiet until wayfinding and programming pull people through. The curator’s version of that problem is a gallery that never fills because the previous room’s exit points the wrong way.

Evidence-led, still human

None of this replaces curatorial judgement. Data does not hang a show. It tells you where the experience the hang intended diverges from the experience visitors had. Calmer, safer galleries — fewer unexplained pinch points — and stories that actually land are outcomes of that feedback loop, not of surveillance.

Because Artty learns from walking habits rather than identities, the visit stays human. Only the patterns come back, aggregate and anonymised. That is what dwell and desire-line data should mean in a cultural institution: a quiet picture of attention and route, good enough to re-hang a room, never detailed enough to profile a visitor.

Practically, teams start with one wing or one outdoor route. They look for rooms with thin dwell relative to their importance in the narrative, objects with clustered dwell that create pinch points, and desire lines that systematically skip a threshold the hang depended on. Then they change one thing — an entrance sightline, a label, a bench orientation — and watch whether the aggregate picture moves. The method is closer to exhibition evaluation than to dashboards for their own sake.

Over a season, those small loops compound. Evidence-led curation is not a single report; it is a habit of treating the visit as readable without treating the visitor as a dossier. Stories that actually land are the ones that survive that habit. Calmer galleries are the ones where desire lines and dwell were allowed to argue with the plan before opening day, and again after.

Related verticals

• Heritage & Museums — Read your galleries like a curator reads a room.
• Waterfronts & Districts — Understand how people use the water's edge.
• DMO & Smart City — See how a whole district actually moves.