Studies | The Fourth Voice

VISITOR STUDIES

What is happening in your gallery

Attendance tells you how many people came. It measures your marketing. It does not tell you whether the exhibition actually worked: what visitors walked past, where they stopped, what held their attention, how long the strongest moments lasted, or whether the element you invested the most in is the one almost nobody notices.

You do not need a master plan to begin answering those questions. Start with three small studies.

THE STUDIES

01

Timing and tracking

Do visitors stop at all? Forty to fifty visitors tracked unobtrusively through one gallery, benchmarked against a published database of 108 exhibitions. Tells you where attention is being won and where it is being lost.

$2,800 · three weeks

02

Label audit

Can they read it when they do? Every interpretive text in one gallery measured against the published Smithsonian standards for length, reading level, type size, contrast and mounting height — and against whether the gallery has one idea its labels serve.

$1,900 · two weeks

03

Language-access audit

Who cannot read it at all? The full visitor journey walked from the website to the front desk to the last label, recording every point at which a visitor who does not read English fluently loses the thread.

$2,400 · three weeks

All three, $6,200, five weeks. Less than the sum of the parts because the fieldwork overlaps, the same visits serve all three studies.

STUDY 01

Timing and tracking

I track 40 to 50 visitors through a single exhibition, one at a time, with a floor plan and a stopwatch: where each person stops, for how long, in what order, and where they leave. No intercepts and no recruiting. Visitors are not aware they are being timed, which is the entire point — cued visitors behave like students, uncued visitors behave like visitors.

Two numbers come out of it.

SWEEP RATE INDEX

The exhibition's square footage divided by the average total time spent in it. How fast people move through your space.

PERCENTAGE OF DILIGENT VISITOR

The share of tracked visitors who stopped at more than half the elements available to them.

These are not numbers I invented. Across Beverly Serrell's database of 108 exhibitions the averages are a sweep rate index of 300 and a diligent-visitor rate of 26 per cent, with a typical visit of about twenty minutes regardless of the exhibition's size or its subject (Serrell, Paying More Attention to Paying Attention). That comparison is the point. It lets you measure your gallery against more than a hundred others instead of against your own expectations.

A written report of 10 to 14 pages, benchmarked against the published record.
An annotated floor plan showing attracting power and holding power at every element, ready to go into a board paper.
A ranked list of elements from most to least visited, with average stop durations.
The dead zones: the specific places where visitors accelerate, skip, or leave early.
The raw anonymised tracking data, so you own it and can re-run the study in three years to measure whether anything changed.

$2,800 for one exhibition or gallery. A second gallery tracked in the same visit, $1,200

STUDY 02

Label audit

Most label problems are not matters of taste. They are measurable, and there are published numbers to measure them against. I assess every interpretive text in one gallery against the Smithsonian standards, then against the harder question underneath: does this gallery have one idea, and do the labels serve it?

Maximum label length of 75 to 100 words; sentences of no more than 25 words, 15 preferred.
Line length of 55 characters at most, with 45 to 50 preferred.
Minimum x-height of 4.5 mm — roughly 24 point — at the closest viewing distance, rising to 48 point at a metre.
70 per cent contrast between text and its background.
Wall labels mounted between 1220 mm and 1675 mm above the floor, with an optimum centreline at 1370 mm.
Exhibition titles of one to seven words; introductory labels of 20 to 125; group and caption labels of 20 to 75.
Eighth-grade reading level or below.

The first five come from the Smithsonian Guidelines for Accessible Exhibition Design; the rest from the Smithsonian Guide to Interpretive Writing for Exhibitions. The reading-level rule is the one institutions resist, so it is worth knowing the Smithsonian's own reasoning: it is behavioural, not intellectual. Visitors who can read college-level prose may simply not want to while on a leisure outing and on their feet.

A label-by-label scoring sheet, with a count of how many fail each standard.
Your gallery's big idea written out in a single sentence — or a plain statement that it does not currently have one, which is the more common finding.
Five labels rewritten in full, as worked examples rather than instructions.
A prioritised fix list that separates what can be corrected in a reprint from what needs a redesign, with the reprints costed roughly so you can act on them this financial year.

$1,900 for up to 50 labels in one gallery. Each additional 25 labels, $600

STUDY 03

Language-access audit

In Florida, 29.8 per cent of residents aged five and over speak a language other than English at home (Migration Policy Institute), against 22 per cent nationally (US Census Bureau). Most institutions interpret entirely in English and have never counted what that costs them.

There is also no standard to audit against, which is why this study is rare. The Smithsonian interpretive writing guide — the most widely used text standard in the sector — contains no guidance at all on translation, bilingual text or language access. The gap is in the literature, not just in your galleries.

The argument worth putting to a board comes from Cecilia Garibay and Steven Yalowitz: would a museum produce an exhibition with no labels at all in the dominant language? Only multilingual labels are routinely asked to prove they increase attendance first (Garibay & Yalowitz, Redefining Multilingualism in Museums). Their scale figures are the ones that end the conversation: more than 60 million US residents speak a non-English language at home, one in five of school age, and 44 per cent of them were born here. This is not a tourism question.

I walk the whole visit — website, booking, car park, arrival, orientation, wayfinding, interpretation, front-of-house scripts, emergency signage — and record every point at which a visitor who does not read English fluently loses the thread. Then I look at whatever translated material you already have and ask whether it was interpreted or merely translated, which are not the same thing and rarely produce the same text.

A journey map marking every break point, ordered by how many visitors it affects.
An assessment of your existing translated text for register, idiom and cultural reference, not only for accuracy.
A language priority grounded in your actual catchment — census tract data for your own postcode, not a national average, because the answer differs sharply between Tampa, Orlando and Miami.
Costed options at three levels: signage and orientation only, core interpretation, or the full journey. Most institutions can afford the first and have never been shown what it would buy.
A one-page language-access policy your board can adopt as it stands.

I work in English and Arabic and assess Arabic text directly. For other languages I can tell you whether a translation reads as interpretation or as a literal rendering, and I bring in a native-speaker reviewer where the text warrants it rather than pretending otherwise.

$2,400 for a single site.

LIMITS

What these will not tell you

These are diagnostic studies. They tell you what visitors did, what your text does, and who your interpretation leaves out. They do not tell you what anyone learned, whether attitudes shifted, or whether behaviour changed after they went home. Attention and access are preconditions for those things, not evidence of them.

Anyone who offers to prove learning outcomes or behaviour change from a study this size is selling you something the method cannot deliver. If those are the questions you need answered, they require a different and larger study, and I will tell you so rather than take the commission.

TERMS

How these run

FEES

Fixed rather than hourly: you know the number before you commit and it does not move. Larger programmes of work are quoted separately.

PAYMENT

A third on signature, a third on completion of fieldwork, the remainder on delivery of the report.

INCLUDED IN EVERY STUDY

A one-hour findings session with your team, in person or remote, and the underlying data in a form you can keep and re-use.

WHAT YOU PROVIDE

A floor plan and permission to be on site during normal opening hours. Nothing else is needed from your staff.

TRAVEL

Included within the Tampa–Orlando corridor. Beyond it, at cost.

ENQUIRIES

The first conversation is free and does not require a decision

Tell me which gallery you are unsure about, and what you suspect is going wrong in it. If none of these three is the thing you need, I will say so and point you at the thing that is.