The right footfall system is the one that answers your question, and no single technology answers them all. Beam counters, cameras, app data, mobile network data and WiFi sensors each count something real and each leave a gap. A hybrid model combines them so the gaps are covered. This issue is a written version of our recent webinar for town centre and retail teams, run with the ATCM and Toolbox Marketing.
Why the data has to hold up
Footfall data has to stand up as evidence, because budgets are tight and every decision needs a justification. Funding programmes such as the Future High Streets Fund and the Shared Prosperity Fund asked for hard data. Footfall gives you the baseline before a change and the proof of what it achieved afterwards. The trouble is that different systems give different answers at the same location, so it matters which one you choose.
Six technologies, compared honestly
Each method has genuine strengths and real limits. Knowing which fits which purpose is the first step to better decisions.
Beam counters and physical sensors
Beam counters are cheap, reliable and good at doorways, but they only tell you how busy a place is. An infrared beam across an entrance logs a count each time it is broken. There is no intelligence behind it: walk through ten times and it counts ten, with no way to tell one visitor from ten. You get no unique visitors, no dwell time and no idea where people went next.
Camera analytics
Cameras count accurately at entrances when they are set up correctly, but they cost more and still count every pass. Modern overhead and 3D stereo cameras use object detection to tell people from trolleys, prams and shadows, which makes them a good fit for store entrances and shopping centre aisles. Each camera sees a limited field of view, so larger spaces need several, and they need maintenance. Like beam counters, they cannot tell you how many individuals came, how long they stayed or whether they came back.
GPS and app data
App data is cheap and covers wide areas, but it is too thin and too modelled for operational decisions, funding bids or lease negotiations. It relies on apps sharing location in the background, which most people turn off. We spent nearly two years testing app data bought from several suppliers against our own sensors, and it captured between 5 and 12 percent of the footfall our sensors saw. The number of signal providers also swung from month to month, so the figures you see are largely modelled. It is useful for high-level trends only.
Mobile network data
Mobile network data gives consistent trends and strong demographics across large areas, but it is not built for granular analysis. Operators supply aggregated data for a nominated area, such as a shopping centre, town or BID area, with age, affluence and home and work origin profiles. The capture rate is around 33 percent, which is low but consistent, so the trends are reliable. It works less well in small areas, can miss short visits and under-18s, and often shows only who was inside the area and for how long, not the routes they took.
WiFi tracking
WiFi tracking can give the most granular picture of how people move, but quality depends heavily on the supplier. Phones with WiFi switched on regularly send out probe requests looking for a network. A good system picks these up, hashes the identifier and uses it to measure unique visitors, dwell time, repeat visits and journeys. A weaker one only sees people who log on to a WiFi network.
That last limit is how GEO-Sense started. We ran a town-wide public WiFi network in Hinckley and tried using its access points to count footfall. It placed people well, but only counted those who logged on, so the numbers were too low. We built our own sensor to fix it.
How GEO-Sense works
GEO-Sense is a footfall sensor we design, build and manage ourselves, and it sees every phone with WiFi switched on, not only phones that connect. We have developed it over ten years specifically to count footfall at the exact points a location needs. Each phone’s identifier is hashed to comply with GDPR, so the reporting shows counts and trends, not who people are.
Because the same hashed identifier can be seen again, GEO-Sense goes further than a counter:
- Unique visitors and repeat visits, so you know how many different people came, not just how many passes were made.
- Dwell time, showing which areas hold people and which they walk straight through.
- Journeys between sensors, with direction and time taken, so you can see how people move around a location and where they enter and leave.
- Live data, in a portal that gives the basics at a glance and detailed analysis when you need it.
Two data sets from one sensor
GEO-Sense reports two measures side by side: unique visitors, and activity (every detection, which is how busy the place is). Comparing them shows things neither shows alone. At one counted location, the two lines tracked each other closely through the morning and early afternoon. From 16:00 they split: activity climbed to around 1,050 an hour by 19:00 while unique visitors held at about 500. The same people were staying and moving around, not more people arriving. A beam counter or camera would have reported that evening as a doubling of visitors.
This is what makes the impact of a change measurable. An event that brings new people in looks different from one that keeps existing visitors for longer, and you need both numbers to tell them apart.
What GEO-Sense cannot see
GEO-Sense will not count people whose phone has WiFi switched off, or people without a phone, including many younger children. That is why we manage the system for you and read the trends with that in mind, and why we recommend combining it with other sources where the decision depends on it.
Why a hybrid model is the strongest option
A hybrid model combines sources so that each one covers what the others miss, giving accuracy, behaviour and consistency in one picture.
- Accuracy: measured counts at fixed points from sensors such as GEO-Sense or cameras.
- Behaviour: how people move, how long they stay and where they come from, from sensor journeys and mobile network data.
- Consistency: calibrated data over years, fit for reporting and funding bids.
More data sets also help calibrate each other. Mobile network data adds demographics and catchment that hardware cannot, while sensors add the granular movement and dwell that network data cannot. Other sources sharpen the picture further: car park transactions from Parking, weather alongside footfall, and sales or retail spend brought in next to the counts. The trade-off is cost, so a hybrid model is best for organisations whose decisions carry real money: councils, retail property owners, BIDs and regeneration programmes.
For most locations, GEO-Sense is the practical backbone of a hybrid model. It already gives you two data sets, works at the granular level other sources cannot reach, and can grow one sensor at a time.
Matching the method to your question
There is no single system that serves every context, so start from the question you need answered. A high street needs a different approach from an enclosed shopping centre. A BID proving its value to levy payers needs something different again, and events, regeneration projects and funding bids each have their own requirements.
Four mistakes to avoid
The costliest footfall mistakes are strategic, not technical.
- Buying on price alone. The cheapest option rarely gives the accuracy needed for funding bids, lease negotiations or BID renewals, and ongoing costs matter as much as the purchase price.
- Relying on a single source. Using one technology for every decision, especially for both day-to-day operations and strategic reporting, leaves blind spots.
- Believing the hype. Find out exactly what data you will receive, where it comes from and what the capture rate is, and check it is enough to support your decisions.
- Not planning for the long term. Footfall is most valuable as a record over years. Short deployments with no continuity make year-on-year comparison, and funding accountability, impossible.
Six tests for any footfall system
Before committing to any system, check it against these six tests.
- Accuracy: does it count correctly, what is the capture rate, and do you know where the data comes from?
- Consistency: does it give reliable trends over time and across conditions?
- Behavioural insight: does it show how people move, how long they stay and where they come from, not just that they were there?
- Scalability: can it grow across locations, zones and time, so you can start with what the budget allows and add to it?
- Accessibility: is the data easy to view, share and understand?
- Cost-effectiveness: does the value of the insight justify the investment at the scale you need?
Talk it through with us
If you want an honest view of your current footfall setup, what is working, what is missing and what we would recommend, book a demo or get in touch. See how GEO-Sense reporting works on the Footfall Analytics page, or how it is used by local authorities, retailers and shopping centres.