Executive summary
SoyXpert, maker of the SoyKey tofu brand in Sherbrooke, Quebec, had no reliable way to measure booth performance at B2B trade shows. Owner Alexandre Gauvin mounted a Storetraffic 3D Scope II LC people counter 13 feet above his booth on a custom collapsible pole, then engineered a multi-threshold workaround to approximate dwell time from an entry/exit-only sensor. The result: a real stop rate and dwell rate he now uses in distributor pitches instead of vague claims about how “busy” a show was.
Key results
- More than half of passersby stopped at the booth at a given show, replacing gut feeling with a measured stop rate
- A large share of those who stopped stayed long enough for a real conversation, not just a glance
- Engineered a four-threshold dwell-time workaround (10 seconds, 1 minute, 2 minutes, 3 minutes) from a single entry/exit counting point
- Replaced a failed DIY sensor build with a single-camera, internet-only setup requiring no fixed wiring
Customer profile
SoyXpert is a tofu manufacturer based in Sherbrooke, Quebec, that sells its SoyKey brand through distributors and exhibits at B2B trade shows to win new retail placements. Owner Alexandre Gauvin relies on trade show performance data to decide which shows are worth returning to and to build the case for SoyKey with prospective distributors.
What challenge did the customer face?
At most B2B trade shows, exhibitors have no real way of knowing how many people walked past their booth versus how many actually stopped. Most never find that out: they pack up, tell their sales manager “we were busy,” and book the same show again next year on gut feeling alone.
For SoyXpert, that mattered because booth staffing and travel both cost money, and choosing which shows to return to the following year is a real budget decision. Without data, that decision came down to memory and vibes rather than something Gauvin could show a distributor or a sales manager.
Why was the previous method insufficient?
Gauvin’s first attempt wasn’t a commercial counting system at all – it was a DIY rig, sourcing parts from Alibaba and using ChatGPT to write the tracking code, on the assumption it would be faster and cheaper than buying one.
It wasn’t. Hauling a full computer setup to a trade show booth, keeping it powered, and trusting it not to fail mid-show was not a serious option for someone who needed to focus on selling tofu, not debugging hardware.
What did SMS Storetraffic implement?
While researching which sensor or reader to buy for his own DIY build, Gauvin’s search led him to Storetraffic instead. He also compared other commercial people-counting systems on the market; most relied on cellular or Bluetooth connectivity, came bundled with more complexity than he needed, and cost more than made sense for a temporary booth setup.
The 3D Scope II LC stood out because it only needed an internet connection: no fixed wiring, no permanent fixture, just a camera and a network link he could bring himself.
How was the rollout handled?
Most trade show venues cap booth height at 8 feet and restrict placing anything directly in front of the booth space, which ruled out a standard booth-mounted camera for the overhead view Gauvin needed. His answer was a custom pole rig, 13 feet tall, built in four collapsible pieces with a telescoping arm so it could pack down and travel between shows.
Show organizers weren’t immediately comfortable with a camera pointed at the crowd, and asked what Gauvin planned to do with the footage before approving the setup. His answer was straightforward: the 3D Scope II LC does spatial counting, not identification. It can distinguish an adult from a child by height, but it isn’t capturing faces or tracking individuals. Once organizers understood that distinction, they signed off – and Storetraffic’s own team confirmed this was the first time a customer had used the product mounted this way at a trade show.
The 3D Scope II LC tracks entries and exits at a single point, which tells you how many people crossed a line but not how long they lingered once inside, so Gauvin had to engineer a workaround for dwell time – arguably the more useful number for a booth, since a visitor who walks past in two seconds is a very different data point from one who stays for three minutes talking to a rep.
His workaround was to layer multiple detection zones at the same physical entry and exit point, each with a different delay threshold: 10 seconds, 1 minute, 2 minutes, and 3 minutes. By exporting the counts from each threshold and subtracting one from the next, he could approximate how long groups of visitors actually stayed. It was an engineered approximation rather than a built-in feature, and it worked well enough to produce numbers he trusted.
What changed after deployment?
The output of that workaround was two numbers Gauvin didn’t have before: at one show, more than half of everyone who walked by the booth stopped, and a large share of those who stopped stayed long enough for a real conversation rather than a glance.
He keeps the exact figures for his own sales deck, but the shape of the result is the point. The improved data reliability allowed him to:
- Show a real stop rate and dwell rate tied to a specific event, instead of telling a distributor SoyXpert “did a lot of trade shows”
- Separate a claim from a story with a hard number, even a rounded one, where most distributor pitches rely on soft language
- Treat trade show data the same way a retailer treats in-store foot traffic data: as evidence a partner can actually check against
As a result, trade show performance became something Gauvin could demonstrate rather than describe.
What can similar exhibitors learn from this?
Gauvin’s setup started with a camera and a pole, not a big IT project, which is worth remembering if your own booth data still lives in memory rather than numbers.
Since deploying the system, Gauvin has found other uses for the same approach: pairing reliable, trained staff with the same traffic and dwell data at in-store tasting demonstrations, so results depend less on any one person’s rapport and more on numbers he can check; and importing sales figures to eventually correlate foot traffic directly with lift. The same camera setup also surfaced a separate problem at a B2B agricultural trade show, where his sales reps were spending so long with each visitor who did stop that they were missing much of the crowd walking past – a gap worth knowing before deciding how a sales team’s time gets spent at the next show. He’s also rethinking sampling strategy broadly: giving away a physical portion of tofu is cheaper than a paper coupon once redemption fees are factored in, and it prompts action right away instead of a coupon that often goes unused.
Why this matters for similar exhibitors
For any business running a temporary footprint – a trade show booth, a pop-up, a seasonal kiosk – a camera, an internet connection, and a bit of engineering can turn “we were busy” into a number a distributor can actually act on. Gauvin has even suggested that convention centers and exhibition halls could install these systems permanently and sell the traffic data back to exhibitors as part of a booth package. Whether or not venues pick that up, the underlying lesson holds: reliable traffic data is what separates a claim from evidence, whether the space is a retail store or a trade show floor.
If you’re weighing whether a system like this would work at your next show, Storetraffic offers a free demo to walk through it.
