People counting accuracy: what the numbers mean by method
How accurate people counting is depends on the method, the placement and the calibration, so no single percentage fits every venue. At a doorway, a 3D sensor or AI camera gives exact in and out counts, while Wi-Fi covers dwell and paths across a whole space. Every deployment is checked against manual control measurements.
On this page
Buyers usually ask about people counting accuracy first, and suppliers often answer with one percentage. A figure like that says little until you know where it was measured, how the sensors were placed and what was being counted.
How accurate is each people counting method?
“Accuracy” means something different at a doorway than across a floor, and each method is strongest at a different kind of count.
| Method | Strongest at | Where it is not the answer |
|---|---|---|
| 3D sensor | Exact directional in and out counts, and occupancy, at a doorway | Behaviour across a whole space |
| AI camera | Point counting with standard camera hardware, analytics only, no personal images stored in counting mode | Whole-space paths and journeys |
| LiDAR | Door and outdoor counting, image-free line detection | Whole-space coverage and journeys |
| Wi-Fi | Whole-space coverage: dwell, paths and zone correlation across an entire store, mall or district | Exact line-crossing in and out at a single door |
| Cellular | Visitor numbers and catchment across city centres, districts and large outdoor areas | Exact counts at a door or inside a single store |
What does an accuracy percentage prove?
On its own, a percentage proves very little. It becomes useful once you know the venue, the placement and the calibration behind it; without those details, you are looking at a figure measured somewhere other than your site. For a buyer, the test is whether the count is accurate and consistent enough for the decision it will feed.
How does Bumbee Labs verify people counting accuracy?
We verify every deployment with manual control measurements matched to our own data, and we tune the model to the space so the numbers match what happened on site. The same verification applies to every method we operate.
Bumbee Labs is the company behind the only footfall method in Europe approved by a data protection authority. The authority is IMY, the Swedish Authority for Privacy Protection.
What should you ask a supplier about accuracy?
- Which authority has reviewed the method? “Compliant” is usually the supplier’s own conclusion. A review by a named authority gives you outside evidence.
- What leaves the sensor? Counts and aggregated statistics, or identifiers and footage?
- How is the deployment calibrated? Are manual control measurements matched to the system’s own data?
- What does the quoted figure assume? Ask for the venue, placement, traffic mix and season behind it.
Which accuracy does your decision need?
A capacity limit at a turnstile needs point precision at that door. A refit decision depends on dwell and paths across the whole space, and staffing is planned from the hourly curve. An exact door count will not show which zone lost visitors after a refit. Many venues therefore combine methods, with sensors at the doors and Wi-Fi across the floor, and use the sensor counts as ground truth for the Wi-Fi model. How to count visitors compares the methods in more detail.
Getting started
Book a demo to see accuracy checked against manual control counts on a venue like yours.
Curious what this looks like for your venue?
Frequently asked questions
How accurate is people counting?
That depends on where and how the count is made. At a doorway, 3D sensors and AI cameras give exact in and out counts. Across a floor, Wi-Fi covers dwell and paths that a doorway sensor never sees. Each Bumbee Labs deployment is calibrated against manual control measurements, so the numbers agree with what happened on site.
Which people counting method is most accurate?
Each method is most accurate at a different job. Sensors and cameras are strongest at point counts, Wi-Fi at behaviour across a whole space, LiDAR at image-free line detection and cellular network data at district scale. Many venues combine methods and let each cover what it measures best.
How is people counting accuracy verified?
Each deployment is checked with manual control measurements that are matched to the system's own data, and the model is tuned to the space. A supplier that cannot describe its calibration procedure is probably quoting a figure from somewhere other than your venue.