Services
Events & Experiences
Enterprise & Training
Gaming & Education
Most installations are judged on how impressive they look. This is a practical guide to judging them on what they actually achieved, the fifteen things worth measuring, the benchmarks that put your numbers in context, and how to decide what success means before anything gets built.
.png)
.png)
This is not an argument against spectacle. Spectacle earns attention, and attention is a prerequisite for everything else. It is an argument against stopping there. A space that impresses everyone and moves nothing is an expensive piece of decoration, and after enough projects you learn to spot the difference well before opening night.
What follows is how we think about measurement: what is worth counting, what the credible benchmarks look like, what to measure depending on what the experience is for, and why almost all of this has to be decided before the technology is specified rather than after it is installed.
QUICK ANSWER: Visual impact is a means, not an outcome. An experience should be judged on what it changed in the visitor, what they understood, felt, did, remembered, or asked for next, because those are the things that connect back to the reason the project was funded.
Every client has a reason for building. A government entity wants a complex national story understood. A developer wants a destination sold before it exists. A brand wants to be remembered in a market where attention is expensive. A museum wants learning. None of those reasons is "a room that photographs well," yet that is frequently the only standard applied once the doors open.
The gap matters commercially. If the only evidence of success is that people said it looked good, there is nothing to defend the next budget with, nothing to optimise, and no way to tell which parts of the experience did the work. Teams end up repeating whatever was most visually dramatic rather than whatever actually performed.
There is a second, quieter cost. Without measurement, no one finds out that the third zone loses half the audience, or that the interactive that took four months to build gets eleven seconds of attention. That information exists. It is simply not being collected.
QUICK ANSWER: Different experiences have different jobs, so success has to be defined at brief stage, not after installation. The definition of success determines the technology specification, the content, the layout and the instrumentation, which is why deciding it late produces a space that cannot be evaluated.
This is the single most important point in this article. Measurement is not a reporting exercise bolted on at the end. It is a design input.
If the objective is qualified leads, the journey needs a natural, non-intrusive data capture moment and the content needs to qualify as it engages. If the objective is learning, the experience needs comprehension checkpoints and content sequenced so understanding builds. If the objective is recall, the design needs a distinct, emotionally weighted peak moment rather than uniformly impressive zones. If the objective is throughput, the layout and dwell targets per zone are engineering constraints, not preferences.
Decide the objective late and you get a space that may well be excellent, but that cannot prove it, and that was optimised for nothing in particular. Decide it early and every subsequent decision has a reference point, including the difficult ones about where the budget goes.
QUICK ANSWER: Useful measures fall into four groups: attention (dwell time, participants, repeat interactions, visitor flow), comprehension (content explored, information retained, learning outcomes), affinity (brand recall, emotional response, social sharing, visitor feedback, return visits) and commercial (leads generated, product enquiries, sales conversations). Most projects should pick three or four, not all fifteen.
Measure card - Pick three or four, not fifteen. Instrumenting everything produces dashboards nobody reads and arguments about which number matters. Choose the two or three that map directly to why the project was funded, plus one diagnostic measure (usually visitor flow) that tells you where to fix things.
Measure card - Separate the diagnostic from the verdict. Visitor flow and content explored are diagnostics: they tell you what to change. Recall, learning and enquiries are verdicts: they tell you whether it worked. Reporting them as one list confuses the conversation.
Measure card - Some of the best measures are not automatic. Recall, retention and emotional response need short surveys at 24 hours and 30 days, not sensors. They are cheap to run and consistently more persuasive to a board than a footfall chart.
.png)
QUICK ANSWER: Dwell time is the most common engagement measure, but the average is close to meaningless on its own. What matters is the distribution, how many visitors reached genuine depth and how dwell compares against the physical footprint the experience occupies.
Some context from published industry measurement. Exposure Analytics' 2026 Physical Engagement Index, drawn from more than 3,000 events and 43.8 million measured visitors across 18 countries since 2020, puts the global engagement rate at around 35.6%, with North America leading at 39.0% and an average dwell of 15 minutes 48 seconds across exhibition environments.
The more useful finding is the spread. Spaces in the top 10% by dwell held engaged visitors for an average of 33 minutes 31 seconds. The bottom 10% averaged 3 minutes 52 seconds. That is close to a ninefold difference between the best and worst performing environments measured the same way. Meanwhile the average dwell at an individual visited exhibition stand sits at roughly 4.2 minutes, down from 5.8 minutes in 2019 — attention is getting harder to hold, not easier.
Three practical implications:
The average hides the answer. A space where most people glance for twenty seconds and a handful stay twenty minutes can report the same average as one where everybody stays four minutes. Those are completely different outcomes. Look at the distribution curve and ask what proportion of visitors reached the depth at which your message actually lands.
Compare dwell against footprint. "Holding power" dwell time relative to the physical size of the exhibit, lets you compare a compact interactive against a large installation honestly. It is often how you discover that the cheapest element in the room is outperforming the centrepiece per square metre.
More dwell is not automatically better. A queue produces dwell. So does confusion. If people are standing still because they cannot work out what to do, that is a design failure reported as engagement. Always read dwell alongside completion and flow.
QUICK ANSWER: A museum should measure learning; an exhibition stand should measure engagement quality and qualified leads; a brand activation should measure recall and participation; an experience centre should measure how effectively complex information was communicated. Applying one metric set across all four is how projects get judged unfairly.
The objective is understanding, not throughput. Time at an exhibit remains the most common proxy for learning in visitor research, but it is a proxy — worth pairing with short comprehension checks and observed behaviour, such as whether visitors explain the exhibit to each other. Note that even highly engaging individual exhibits typically hold people for only three to four minutes, so targets should be set against that reality rather than against hope.
Engagement quality and qualified leads, not badge scans. Visitor-to-lead conversion in the range of 10–30% is a far more honest measure of stand performance than raw scan volume. For cost context, industry research puts a trade show lead at roughly USD 142 against roughly USD 596 for a field sales call — which is the actual argument for exhibiting, provided the leads are followed up.
Recall and participation. Published benchmarks suggest 65–80% of consumers recall a brand experience they participated in after 30 days, against 10–25% for digital advertising, with 30-day recall of 50–60% considered good, 60–75% strong and above 75% exceptional. If you are running an activation and not measuring recall at 30 days, you are leaving the most flattering and most decision-useful number on the table.
How effectively complicated information was communicated. This is the hardest thing on this list to measure and the most valuable. The practical approach is a short structured debrief with hosted visitors: what did they take away, what could they now explain to a colleague, what did they ask about afterwards. For B2B experience centres, the honest commercial measure is whether the visit changed the shape of the sales conversation that followed.
QUICK ANSWER: Experiential outperforms passive media on recall because emotionally charged, multisensory experiences are encoded more strongly and retrieved through more pathways. Published work suggests such experiences can be retained several times longer than single-channel information delivery.
This matters for measurement because it tells you what to design for, not just what to count.
Memory formation is strongly linked to emotional response, emotionally weighted material is encoded more durably than neutral material. Multisensory experiences compound the effect by creating multiple retrieval routes: sight, sound, touch and movement each provide a separate path back to the same memory. Research in this area indicates emotionally charged multisensory experiences can be retained substantially longer than single-channel delivery, with some studies putting the difference at around six times.
The practical design consequence is that memorability is not evenly distributed across a visit. It concentrates around a small number of emotionally distinct moments. This is why we argue so consistently for two or three genuine peaks rather than uniform intensity: peaks are what get encoded, and encoded moments are what your 30-day recall survey will actually capture.
It also explains a common measurement surprise. The zone that scores highest on dwell time is frequently not the zone visitors describe when asked a month later. Both numbers are correct. They are measuring different things.
QUICK ANSWER: The common failures are measuring only after launch, counting reach instead of depth, collecting data nobody acts on, and, most expensively, generating leads that are never followed up.
Instrumenting after the fact. Sensors, analytics and survey mechanisms are far cheaper to design in than to retrofit. Deciding to measure after opening usually means measuring whatever is easy rather than whatever matters.
Counting reach instead of depth. Footfall and impressions are the easiest numbers to produce and the least informative. Industry reporting has been moving away from them for exactly this reason: engagement rate, dwell depth and lead relevance tell you something actionable; a footfall total rarely does.
Letting the leads die. This is the most costly mistake in the entire field. Industry analysis indicates up to 80% of trade show leads never receive any follow-up at all, and that companies following up within 24 hours are several times more likely to convert than those waiting a week. No installation, at any budget, can compensate for that.
Collecting data no one owns. If nobody is accountable for reviewing the numbers at 30, 90 and 180 days, the measurement will quietly stop. Name the owner at brief stage.
Measuring once. A single post-launch report tells you how the opening went. Reviewing at intervals tells you how the asset is performing, which is the question that matters for anything permanent.
.png)
QUICK ANSWER: Write the success definition into the brief, choose three or four measures, instrument during design, capture a baseline, and schedule reviews at 30, 90 and 180 days with a named owner.
Step 1, Write success into the brief. One or two sentences stating what the experience must achieve and for whom. If it cannot be written plainly, the project is not ready to be designed.
Step 2, Choose three or four measures, and say what good looks like. A target you would be satisfied with, and a number you would consider a failure. Both, in writing, before anyone falls in love with the results.
Step 3, Instrument during technology design. Sensors, interaction logging, flow tracking and any data capture points get specified alongside the AV and show control, not afterwards.
Step 4, Capture a baseline. Where possible, measure the existing situation, current enquiry volume, current comprehension, current dwell in the space being replaced. Without a baseline you will have numbers but no way to argue they represent improvement.
Step 5, Schedule the reviews and name the owner. 30, 90 and 180 days, with one person accountable and a standing decision to act on what the data says, including reworking content that underperforms.
Step 6, Budget for the second version. The first release of any experience is a hypothesis. The projects that perform over years are the ones where a modest content and optimisation budget was reserved from the start.
QUICK ANSWER: We treat the definition of success as part of the design brief rather than a reporting task, specify instrumentation alongside the technology, and review performance after launch. Across more than 800 delivered projects, this is the clearest difference we see between experiences that keep performing and ones that peak on opening night.
We are a technology company, and it would be convenient for us if the answer to every brief were more technology. It usually is not. The most useful conversation we have with a client is often the one where we establish that the objective can be met with a smaller, sharper intervention than the one originally imagined, and then agree how we will both know whether it worked.
Success is defined at brief stage, with the client. Before specification, we agree what the experience has to achieve, for which audience, and which three or four measures will be used to judge it. This is written down. It becomes the reference point for every later trade-off.
Instrumentation is designed in, not retrofitted. Interaction logging, flow tracking, dwell measurement and any data capture moments are specified alongside the AV, sensing and show control layers, so the data exists from day one and the capture feels like part of the experience rather than an interruption.
We review after launch, not just at handover. Zone-level dwell, interaction rates, completion and flow get reviewed against the agreed targets, and the findings drive content and pacing changes. Permanent installations are assets, and assets are managed.
The track record is what makes the estimates credible. More than 800 projects delivered since 2018 across the UAE, Saudi Arabia, Oman and the wider GCC, for clients including DEWA, the Government of Dubai, Atlantis The Royal, Emaar, Al Jalila Foundation, PwC, VISA and Louis Vuitton, means the targets we propose at brief stage come from what comparable experiences have actually achieved in this region.
Start by defining what the experience is meant to achieve, then choose three or four measures that map to it, typically a mix of attention (dwell time, participation), comprehension (content explored, information retained), affinity (recall, return visits) and commercial (leads, enquiries, sales conversations). Instrument those measures during design, capture a baseline beforehand, and review at 30, 90 and 180 days against targets agreed in advance.
It depends entirely on format, and the average matters less than the distribution. For context, published 2026 benchmarks put average exhibition dwell around 15–17 minutes overall, while an individual trade show stand averages roughly 4.2 minutes and individual museum exhibits often hold visitors only three to four minutes. The more useful question is what proportion of your audience reached the depth at which your message lands.
Only as context. Footfall tells you how many people were nearby, not whether the experience worked. Engagement rate, the share of passers-by who actually interacted, and dwell depth are far more informative, which is why industry measurement has been moving away from reach metrics toward engagement quality.
Thirty days is the standard and the most decision-useful window. Published benchmarks suggest 65–80% of participants recall a brand experience at 30 days versus 10–25% for digital advertising, with 50–60% considered good, 60–75% strong and 75%+ exceptional.
No. An event is measured once, against a fixed objective, and the data informs the next event. A permanent installation is an asset, so measurement is continuous: performance is reviewed at intervals and used to refine content, pacing and staffing over years, not just reported at handover.
Generating leads and never following them up. Industry analysis suggests up to 80% of trade show leads receive no follow-up, and that responding within 24 hours makes conversion several times more likely. No amount of investment in the experience itself can recover value lost after the visitor walks away.
Usually yes, though less elegantly and at higher cost than designing it in from the start. An audit will typically identify where sensing, interaction logging or a light-touch data capture moment can be added, and a baseline can be established from the current performance before any changes are made.
We design and build immersive experiences that engage, excite and transform audience interaction combining art, engineering, and next-gen technology to bring brands and spaces to life.
