Planning For People In Museum Exhibitions

The Practical Reality of Human-Centered Exhibition Design

Most museum exhibitions fail at the planning stage not because the content is bad but because nobody walked through the space before the walls went up. I learned this the hard way on a mid-size natural history exhibit I helped coordinate back in 2018. We had a beautiful layout on paper. We placed the touch-screen interaction tables along the main corridor at what we thought was eye level. What we missed was that the corridor was a two-way flow path with an average dwell time of under three minutes. The moment the screen loaded, two visitors would stop, block the path, and create a bottleneck that backed up into the adjacent gallery. Within a week of opening, we had a maintenance issue: scratched screens from elbow contact because people were leaning against the table while waiting for the queue to clear. The fix wasn't a software update. It was moving the stations to a side alcove and widening the corridor by eighteen inches using existing display cabinetry. That redesign cost us about forty hours of labor and pushed the opening back ten days. It also eliminated ninety percent of the traffic complaints we received in the first month. The core of this work isn't abstract. It's about predicting how bodies move through space and how attention behaves under real conditions. There's a concept called "traffic flow analysis" that you'll see in exhibition design specs, and it sounds formal but it's basically just watching how people actually walk somewhere before you commit to a floor plan. You set up a camcorder in a corner of the building, record foot traffic for two hours during peak visiting time, and then you draw lines on a printed floor plan showing where people naturally pause, turn, and cluster. I use this method because it's cheaper and faster than buying any of the expensive simulation software that vendors try to sell you. It also gives you data you can argue with in meetings, which matters when a curator insists on putting a forty-foot diorama in the only spot that isn't a choke point. Another thing nobody tells you: accessibility isn't a compliance checklist. It's a design variable that changes everything. A ramp that meets ADA requirements might still force a wheelchair user to back up into traffic. A braille label placed at standard height might be unreadable because the lighting creates glare at that exact angle. I once spent three days repositioning tactile floor indicators because the carpet texture in the corridor was too thick for cane detection, and the smooth tile override we installed created a tripping hazard at the threshold. These aren't theoretical problems. They're the kind of things that show up after opening day when you're fielding complaints instead of making decisions.

What Actually Goes Into the Planning Process

Let's start with the sequence most teams get wrong. You don't begin with the content. You begin with the space. Measure everything. Floor load capacity, ceiling height, HVAC output, structural columns, existing utility conduits, fire egress routes, and the actual width of the delivery door that your freight elevator can reach. A colleague of mine once designed an entire interactive exhibit around a custom steel fabrication that couldn't fit through the service entrance. They ended up disassembling it piece by piece on the loading dock, spending two extra days and eight thousand dollars in labor to reassemble something that should have fit on the first try. The lesson is straightforward: verify your physical constraints before you lock in any fabrication specs. After you map the space, you sequence the visitor journey. This isn't about making a narrative arc. It's about managing cognitive load. The average museum visitor spends roughly forty-five to sixty minutes in a standard gallery before fatigue sets in. After that, attention drops and engagement becomes performative rather than genuine. You'll see it in the data: dwell time on interpretive panels falls by about sixty percent after the first twenty minutes of continuous viewing. So the practical move is to break the experience into segments separated by physical transitions. A change in floor material, a shift in lighting temperature, a doorway that narrows or widens. These micro-pauses reset the visitor's mental state without requiring any explicit instruction. Interactive elements need a different kind of planning because they introduce friction. Every touch screen, audio station, or hands-on display adds a point where someone stops moving. You need to calculate the spatial footprint of that stopped person plus their personal bubble. The standard guideline is twenty-four to thirty inches around a stationary visitor, but in practice you want forty-eight inches minimum because most people will lean or shift when engaged. A single interactive station with insufficient clearance creates a domino effect. Two stations placed sequentially without adequate buffering do the same. I always recommend staggering interactive elements across different zones rather than clustering them, even if the thematic logic suggests grouping them together. Thematically coherent doesn't mean spatially efficient.

Signage and Wayfinding Under Real Conditions

Sign placement follows a different set of rules than you'd expect. The conventional wisdom is to put directional signs at decision points: intersections, doorways, platform edges. That's correct but incomplete. You also need to consider sightlines from standing height versus seated height, because families with young children and wheelchair users will be making those same decisions. Text legibility distance matters too. A standard overhead sign with twelve-point type is readable at about fifteen feet. By thirty feet, most adults are struggling to resolve the characters, and children or visitors with low vision won't read it at all. I use a simple formula: minimum character height in inches equals the viewing distance in feet divided by ten. So a sign meant to be read at forty feet needs four-inch characters minimum. It sounds obvious but I've seen galleries where the wayfinding font was scaled for print brochures, not environmental reading distances. Acoustic planning is another area where teams consistently underinvest. Open galleries with hard surfaces create reverberation times that can exceed three seconds. At that level, spoken audio from interactives becomes unintelligible, and even quiet conversations between companions are effortful. The workaround isn't always expensive. Acoustic panels behind display cases, fabric-wrapped baffles suspended from ceilings, and strategic placement of sound-absorbing materials behind open shelving can bring reverberation down to under one point five seconds without destroying the visual aesthetic. I once worked with a designer who rejected acoustic treatment because she didn't want to "compromise the aesthetic." Six months into the exhibit, we removed three large velvet drapes from the windows and replaced them with the baffles. The visual impact was negligible. The audio clarity improved dramatically.

Lighting and Visual Comfort

Lighting design for exhibitions isn't about making things look good. It's about making things readable without causing discomfort. The standard illuminance range for general gallery circulation is fifty to one hundred lux. Display cases with sensitive materials need lower levels, often twenty to fifty lux depending on the object. The problem is that these requirements conflict with interactivity. Touch screens and projection surfaces need higher ambient light to be visible, but that same light damages light-sensitive objects nearby. The solution is localized task lighting. A focused LED spot at two to three times the ambient level on the screen surface eliminates glare and reflection without raising the overall gallery illumination. I recommend specifying dimmable, color-corrected LEDs with a CRI above ninety because cheap LEDs will shift the color rendering of adjacent artifacts and trigger conservation complaints within months. Gloss and glare are the invisible enemies of exhibition planning. A polished concrete floor with overhead track lighting becomes a mirror that blinds visitors walking toward a display. A backlit panel viewed at a forty-five degree angle from the main approach creates a reflection that blocks the content entirely. The fix is measurement. Take a light meter reading from the visitor's expected viewpoint, not from the ceiling. If the reflected luminance exceeds the emitted luminance of your display surface, you have a glare problem. Adjust the angle of the display, change the finish, or reposition the light source. This takes about five minutes per station and prevents the seasonal complaint wave that hits galleries during winter when the sun angle is low and natural light becomes a factor.

Capacity and Crowd Management

Every exhibition has a maximum comfortable capacity that is almost never the code maximum. Building codes tell you how many people can exit a space in an emergency. They don't tell you how many people can occupy the space before it feels suffocating. I use a rule of thumb: one square meter per person for circulating space, and three square meters per person for stationary engagement areas. A two hundred square meter gallery with mostly circulation space holds about two hundred visitors comfortably. At three hundred, people start avoiding eye contact. At four hundred, the experience degrades noticeably. The numbers matter because ticketing systems rarely account for space usage patterns. A timed entry slot that allows five hundred people into a space designed for two hundred will produce a poor experience regardless of how well anything else is planned. Queue design deserves more attention than it gets. A poorly designed queue creates frustration independent of the exhibit itself. The minimum queue width is thirty-six inches for single-file flow. Forty-eight inches allows two people to pass without collision. A serpentine queue with sharp ninety-degree turns forces visitors to rotate their bodies and slows flow by about thirty percent compared to gentle curves with a radius of at least three feet. I once redesigned a museum queue by replacing a series of sharp corners with a gradual spiral. The same floor area held twenty percent more visitors in line without increasing perceived wait time. The difference was entirely about body mechanics and spatial comfort.

Testing Before Opening

Soft opening is the most underutilized tool in exhibition planning. It's not a marketing event. It's a diagnostic process. Invite a controlled group of visitors for free admission on a weekday morning and give them a simple task: find the exit, locate the restroom, interact with one station, and tell you what confused them. You'll learn more from those fifteen minutes than from any amount of stakeholder review. I track three metrics during soft openings: time to first interaction, incidence of path obstruction, and spontaneous verbal feedback. The verbal feedback is unstructured but highly reliable. Visitors will tell you exactly what they found confusing if you ask them after they've had a chance to move through the space. Written surveys collected at the exit are almost useless because the experience has already faded and people default to polite generalities. There's also value in observing visitor behavior without any prompt. I bring a notebook and sit in a corner for the first hour. I note where people stop, where they hesitate, where they double back, and where they seem frustrated. These observations generate a list of priority fixes that usually runs between five and fifteen items. Some are trivial: a sign tilted at an awkward angle. Some are structural: an interactive that requires two people to operate but is placed in a solo-path location. The stubborn ones are the ones that require renegotiation with stakeholders because they conflict with the curatorial vision or the budget. Those are the ones you need data to defend, which is why the observation period matters.

Where This Approach Breaks Down

Human-centered exhibition planning doesn't work when the institution treats it as a checkbox. If the goal is to produce documentation that satisfies a funder or a board rather than to actually improve the visitor experience, the process becomes performative. You'll get the traffic flow maps and the capacity calculations but nobody will act on them because the decisions were made before the planning started. This is common in repurposed spaces where the building itself dictates constraints that no amount of planning can fully resolve. An old warehouse with low ceilings, narrow doorways, and uneven floors can still produce a compelling exhibit, but the planning needs to embrace those limitations rather than fight them. Attempting to force a modern exhibition layout into a historically constrained building usually results in compromises that satisfy no one. Another scenario where this falls apart is when the budget for physical modifications is nonexistent. The fixes I've described—repositioning displays, adding acoustic treatment, redesigning queues—require labor and materials. If those resources aren't available, you're working with whatever exists. That's not a failure of the planning methodology. It's a reality of institutional constraints. The best you can do in that situation is prioritize the highest-impact adjustments and document what you couldn't address so that the next iteration has a baseline to build from. I keep a running log of every exhibition I plan, noting unresolved issues and their estimated impact. This log becomes the starting point for the next project and tends to reduce repeat mistakes by about half. The one honest limitation I need to state plainly: this planning process cannot predict every edge case. Visitor behavior contains enough variability that some problems will only emerge after opening. The goal isn't to eliminate all surprises. The goal is to eliminate the ones that are obvious in hindsight and to reduce the severity of the ones that aren't. A well-planned exhibit might still have a broken interactive or a confusing sign. But it won't have the systemic failures that make the entire experience feel broken to the people who walked through the doors expecting something better.