Museum dwell time: the exhibits that hold visitors

There is a permanent collection gallery in one of Europe’s most visited museums that curators long assumed was a highlight. It anchors one wing of the building, has a prominent place in the audio guide, and the objects on display are considered the institution’s most significant holdings. When the museum finally measured exhibit dwell time (how long visitors actually spent there), the answer was uncomfortable: the average visitor passed through in under four minutes.

The gallery that followed it, smaller and assembled on leftover budget, showed dwell times more than twice as long.

The two galleries show the gap between what museums believe about their visitors and what those visitors actually do. Measuring that gap is no criticism of the curation; it is a way to understand how a building works.

What dwell time measures

Dwell time is the time a visitor spends in a defined zone. It is easy to calculate for a single zone and becomes informative at scale, across every gallery and corridor in a large institution. Museum researcher Beverly Serrell spent years compiling dwell times across more than a hundred exhibitions in museums, zoos, and aquariums (work published in the journal Curator in 1997) and developed the concept of the “sweep rate index”: how quickly, per square metre, visitors move through a space.

Her finding, still quoted in exhibition design literature, is that most visitors are “casual” rather than “thorough”, and that exhibitions routinely overestimate how long people will spend at any given element.

That is useful to know. Operational decisions need the same knowledge for your specific building, not an average across many institutions.

High dwell at an exhibit signals engagement: reading a label, watching a video, returning with a companion to point something out. Exhibit engagement measured this way is a far more honest performance signal than ticket sales or total visitor counts. Low dwell is not necessarily disengagement (a short corridor linking two galleries is built to be walked through and will show low dwell), but low dwell at a destination exhibit is worth investigating. The interpretation may be too dense, the lighting may be drawing the eye elsewhere, or a bottleneck at the entrance may be giving visitors a reason to keep moving.

Flow: the route that visitors actually take

Dwell tells you where people stop. Flow tells you how they move between stops, including what they miss.

In most museums, visitors do not follow the intended sequence. They enter a gallery from the “wrong” side, skip the introductory panel, and encounter a narrative out of order. Some wings get discovered only by visitors who take a wrong turn. Others are reliably walked through without a single pause.

Visitor analytics for museums and cultural venues maps these actual routes rather than the intended ones. The output describes how real visitors, with real time constraints and real interests, navigate the institution, and that description lets layout changes and interpretation decisions rest on evidence instead of instinct.

Queue and crowding data comes from the same source. A single popular exhibit (a borrowed object, a blockbuster loan, a piece that went viral) can create queues that back up into adjacent galleries and suppress dwell time in spaces that would otherwise perform well. Identifying that dynamic from data, rather than from staff anecdote, means a venue can act on it: adjusted opening hours, a timed slot for the popular item, or a flow intervention that redistributes visitors before the queue forms.

Reporting to funders: a number that has to be right

Most cultural institutions also report their visitor statistics to funders (Arts Council England, local authorities, charitable foundations, government departments) as the primary demonstration of public value.

A five-year analysis by Arts Council England of around 1,200 accredited museums in England found that as of 2023–24, average annual visitor numbers remained 10% below pre-pandemic levels, with half of respondents not yet seeing recovery.

In that environment, the accuracy of reported figures matters even more.

Manual counting (clickers at the door, tally sheets from front-of-house staff) is error-prone and incomplete. It misses re-entries, secondary entrances and visitors on days when the counter wasn’t watching. Automated footfall counting produces a continuous, auditable record that closes those gaps. The data deliverables include daily and hourly counts by entry point, which satisfy funder reporting requirements and provide an accurate baseline for year-on-year comparisons.

Museum and gallery visitor analytics show that the institutions getting most value from this data are those that combine headline visitor counts with zone-level dwell and flow, using the headline for external reporting and the detailed layer for internal decisions.

The exhibit that earns its square metres

Museums have limited floor space. Every square metre that houses a gallery carries a cost (maintenance, climate control, staffing, interpretation), and the implicit question is whether the objects and experience in that space justify it. Dwell time and flow data let you answer that question with some rigour.

An underperforming gallery might need re-interpretation. A high-performing one might benefit from expanded space, or better signage to route more visitors towards it. Temporary exhibitions can be planned around the zones that historically show the strongest engagement as well as the ones with the largest floor area.

Libraries, which share many of the same challenges (publicly funded, reporting to funders, managing a building that serves multiple purposes simultaneously), find similar value in zone-level analytics: which parts of the building people use most, and which sections see the least dwell. The questions are the same; the interventions differ.

What dwell time data can and cannot tell you

Dwell time measures time in a zone. It does not measure what a visitor was thinking, feeling, or learning during that time. A long dwell could be genuine fascination or a visitor sitting down to rest their feet. Short dwell could be a rapid, expert scan by someone who knows exactly what they came to see.

So dwell data works best alongside other inputs (visitor surveys, front-of-house observation, interpretation evaluation) rather than as a replacement for them. It describes behaviour and should not be read as a verdict. The signal is highly repeatable, available continuously, and covers the entire building rather than a sample of exits.

The measurement is anonymous throughout: no visitor is identified and no image is captured. The analytics system produces aggregate flow and dwell statistics that describe the building’s behaviour patterns, not any individual within it. For institutions whose relationship with visitors is built on public trust, that distinction matters.