How do you count visitors?

What question should your visitor count answer?

A visitor count can answer several questions: how many people came in, how long they stayed, which zones drew them, how many passed by without entering and how full the building is right now. Different methods answer them. A visitor counting system that suits one of these uses can be wrong for another, so write down what you need to know first and pick the technology afterwards.

The short version

• Wi-Fi measures behaviour across whole spaces, with no new hardware for most venues.
• 3D sensors, AI cameras and LiDAR give exact counts at a point such as a door, and none of them stores personal images when counting.
• Cellular network data covers districts and outdoor areas, but it gives no indoor detail and no exact door count.
• Many venues combine Wi-Fi, 3D sensors and mobile network data in a hybrid, so each method covers the others’ weaknesses.

How do the visitor counting methods compare?

Can Wi-Fi count visitors across a whole space?

Yes. A wifi people counter turns the enterprise access points a venue already runs (Cisco, Ruckus, Aruba and more) into a counting network. Phones with Wi-Fi switched on give off signals as people move. These signals are anonymised and aggregated into visitor counts, dwell time, paths between zones, passers-by and capture rate, and AI filtering removes noise such as stationary devices and staff. Manual control measurements calibrate every installation against reality, and how counting without cameras works explains the principle.

Most venues need no new hardware. Wi-Fi is weaker at a single doorway, though: for exact line-crossing at one entrance, and for occupancy at a door, 3D sensors are the better tool.

When do exact door counts call for 3D sensors or AI cameras?

3D sensor counting uses overhead optical 3D sensors, such as Xovis and Flir/Brickstream, with AI analytics. Where standard camera hardware suits the site better, an AI people counting camera does the counting inside the camera, which makes it practical where camera infrastructure already exists. Both are built for places where a number has to be exact: directional in/out at an entrance, occupancy against a safe limit, and queues and dwell at a bottleneck. In counting mode no personal images are stored.

Their limit is reach. A sensor is precise at its point but registers nothing beyond it, and 3D sensors mean installing hardware at every point you want to count.

Does LiDAR count visitors without a camera?

Yes. LiDAR people counting works from eye-safe laser pulses that sweep the space. People show up as anonymous clusters of points, never as pictures, so no camera is involved and no image exists. The result is exact counts, occupancy and paths, indoors or outdoors and in any light, and the method is strongest where lighting is difficult or the area is wide open. If you are choosing between LiDAR and a camera, read LiDAR vs camera people counting.

Cellular: counting at district scale

Cellular (4G/5G) network analytics covers whole districts, streets and open outdoor areas. It works from anonymised, aggregated mobile network data and answers the wide questions: total footfall along a high street and which districts visitors come from, per day and season. Indoor detail and an exact count at a door are outside its reach.

Why do many venues combine methods?

Each method has a blind spot, so many venues run a hybrid setup: Wi-Fi for behaviour across the space, 3D sensors at the doors that need exact counts, and mobile network data for large outdoor areas. The streams cross-validate each other, and sensor ground truth calibrates the Wi-Fi model, which raises confidence in every number.

Which method fits which venue?

• A single entrance where the count must be exact: a 3D sensor. If you are weighing it against a basic door counter, read people counter vs people counting analytics.
• Stores, showrooms and pop-ups: Wi-Fi for dwell, zones and capture rate, plus a sensor at the door where precise in/out matters.
• Shopping centres, transport hubs and large venues: Wi-Fi across the whole space, with sensors at the main entrances.
• Museums and galleries: Wi-Fi across the galleries and a sensor at the entrance; see visitor analytics for museums.
• Airports: a hybrid, from kerb to gate.
• City centres, districts and outdoor areas: cellular for the wide view, Wi-Fi where the detail matters.

How does a count become statistics?

Whatever mix you choose, the data lands in one analytics platform with a people counting API. Dashboards give each stakeholder their own view, and an open, documented API feeds the BI tools you already run. Standard metrics are ready the day after the visit, and recurring visits within the same day can also be measured. Where occupancy matters, near-real-time crowd alerts come with the Crowd Alert add-on. The data deliverables page lists which method delivers which metric.

With every method, the data you receive is anonymous and aggregated. The compliance detail is on GDPR-compliant footfall analytics. Bumbee Labs is the company behind the only footfall method in Europe approved by a data protection authority.

Accuracy, method by method, is covered in people counting accuracy by method. All resources and guides collects the guides, the glossary and selected insights; all insights are under insights.