Footfall analytics for shopping centres: WiFi, cameras and sensors compared

Shopping centres measure footfall best by combining entrance counters, such as stereo cameras or lidar, with cloud WiFi analytics on existing access points from Cisco Meraki, HPE Aruba, or Ruckus. Entrance counters deliver headline counts, while ISO 27001 certified WiFi analytics tracks dwell and zone movement across 80,000+ venues without requiring new hardware.

Which footfall measurement method should your shopping centre choose?

When you buy footfall analytics for shopping centres, you are buying answers to three separate questions:

• How many people came in?
• Where did they go, and for how long?
• Where did they come from?

No single technology answers all three well, so the choice is rarely either-or.

• Entrance counters answer the first question. Stereo cameras, thermal sensors and lidar count people crossing a line, in both directions.
• WiFi presence analytics answers the second. Your access points detect phones across every mall, food court and car park they cover.
• Mobile data panels answer the third. They model visits from a sample of phones and show where visitors live and work.

Our recommendation follows the order of those questions:

1. No reliable door count today? Install entrance counters at your main public entrances first. Valuations, service charge discussions and investor reports rest on that figure.
2. Already counting your doors? Add WiFi presence analytics before you buy more sensors. It extends measurement into zones and tenants for the cost of software.
3. Need catchment, acquisition or competitor comparison? Buy panel data.

Where do WiFi, cameras, sensors and mobile data genuinely differ?

Four things separate these options for an asset or centre manager: accuracy, zone coverage, privacy exposure and cost. Cost follows from the first three, so start there.

Accuracy: people at a line versus devices in a zone

Entrance counters measure people.

• Stereo cameras use two lenses to judge height. That lets them filter out trolleys and pushchairs and separate adults from children.
• Thermal sensors count body heat crossing a line.
• Lidar fires laser pulses to build a 3D point cloud. It follows each shape across a wider area.

All three depend on correct mounting and regular validation against a manual count. A sensor knocked out of alignment during a shopfront refit can undercount for months unnoticed.

WiFi presence analytics measures devices, not people. Some shoppers carry no phone, some carry two, and some switch WiFi off. Apple and Google both document that modern phones randomise the hardware (MAC) address they broadcast. That makes phones that never connect hard to deduplicate across a single visit. WiFi is therefore strong on trends, dwell and zone share, and weak on exact headcount.

Connected visitors give a far cleaner signal. Some visitors sign in through a captive portal, the sign-in page shown before network access. Each of those visitors creates one consented record per visit. You can then measure their dwell and return frequency with confidence.

The practical fix for the device-versus-person gap is calibration:

1. Run WiFi alongside your door count for four to eight weeks.
2. Calculate the ratio of detected devices to counted entries.
3. Apply that ratio to the doors you do not count.
4. Recheck it every quarter.

Coverage: entrances versus zones and tenants

A door counter stops at the door. Take a centre with 12 entrances and 150 units. Its counters show how many people came in, not which malls they used. Covering zones with sensors means another counting line at every mall junction and tenant entrance. Hardware, cabling and maintenance costs rise with every line you add.

WiFi covers wherever your access points do: malls, food court, leisure and car park lobbies. That makes it the practical route to location analytics in shopping centres. You can report:

• dwell time by zone
• movement between anchor stores
• the share of visits reaching an upper level

One access point often serves several units. Report by zone or mall rather than by individual store.

Privacy and UK GDPR

Each method carries a different level of exposure under UK GDPR and the Data Protection Act 2018.

• WiFi presence analytics. The ICO's guidance on WiFi location analytics treats device identifiers such as MAC addresses as personal data where they can single out an individual. Hashing pseudonymises identifiers but does not anonymise them (UK GDPR Recital 26). Run a Data Protection Impact Assessment (DPIA) under Article 35. Display signage at entrances and explain how visitors can opt out.
• Stereo cameras. If the device processes images of identifiable people, UK GDPR applies in full, as does the ICO's video surveillance guidance. Ask whether counting happens on the sensor and whether any image ever leaves it.
• Thermal and lidar. Neither captures a recognisable face, so the risk is lower. Document the processing in your DPIA anyway, because visitors cannot tell a sensor from a camera.
• Mobile data panels. The panel supplier is the controller for the source data, and you receive aggregates. Ask how panel members gave consent and what minimum group size the supplier reports.

You may also use WiFi sign-ins for email marketing. If so, the Privacy and Electronic Communications Regulations (PECR) require separate consent. Purple's conscious-choice opt-ins keep marketing consent apart from network access.

How the five methods compare

Criterion	WiFi presence analytics	Stereo entrance cameras	Thermal sensors	Lidar sensors	Mobile data panels
What it counts	Phones detected or connected on your network	People crossing a door line	People crossing a line, by heat signature	People and their paths across an open area	Modelled visits from a phone panel sample
Best output	Dwell, zone flow, return visits	In and out counts per entrance	In and out counts in low light	Counts and flow across wide entrances and atria	Catchment, origin, cross-site comparison
Zone and tenant coverage	Every area with access point coverage	Doorway only	Doorway or narrow corridor only	One defined open zone per sensor	Whole site, little detail inside the building
New hardware	None where access points already cover the site	One or more ceiling units per doorway	One or more ceiling units per doorway	One or more high-mounted units per zone	None
Main accuracy limit	Counts devices, not people; MAC randomisation	Mounting height and door geometry	Background heat near open doors	Line of sight and occlusion	Sample size and modelling
UK GDPR exposure	Personal data; DPIA, signage and opt-out	Personal data if images are processed	Low; no identifiable image	Low; point clouds hold no faces	Supplier is controller; you receive aggregates
Who it suits	Asset and marketing teams needing zone and repeat-visit data	Centre managers reporting headline footfall	Entrances with poor light or privacy concerns	Open-air schemes and wide entrances	Asset managers comparing catchments

When are entrance cameras or sensors the right call?

Choose entrance counters when the number itself carries commercial weight. That applies in four situations:

• You report footfall to investors, lenders or valuers, who expect a people count from a named method.
• You benchmark against national series built on sensor counts, such as the BRC-Sensormatic IQ Footfall Monitor.
• Your entrances are defined doorways, where a single counting line captures everyone.
• You need in-and-out flow per entrance to plan cleaning, security and car park staffing.

Stereo, thermal or lidar: which entrance sensor fits?

Stereo cameras suit standard indoor mall entrances. Height filtering separates adults from children and trolleys. On-device processing keeps images out of your systems.

Thermal sensors suit entrances in low light, and sites where visible cameras would concern visitors or tenants. They need care at doors facing direct sun or overhead heaters, which shift the background temperature.

Lidar suits wide open entrances, atria and open-air schemes. One unit covers a larger area and follows paths, not just line crossings. It needs higher mounting points and a clear line of sight.

When is WiFi presence analytics the right call?

Choose WiFi presence analytics when your question is about behaviour inside the centre rather than the count at the door. It fits when:

• your malls, food court and car park already have WiFi coverage
• you need dwell time, zone flow and return-visit frequency for leasing, marketing or tenant mix decisions
• your scheme has no doors to count, such as a retail park with open frontages
• you want first-party data from consented sign-ins, so you can reach visitors after they leave

Purple's WiFi Analytics runs as a cloud overlay on the access points you already own. It is hardware-agnostic and works with Cisco Meraki, HPE Aruba, Ruckus, Juniper Mist, Ubiquiti UniFi, Cambium, Extreme and Fortinet. No rip and replace is required. See why your next WiFi upgrade doesn't need new hardware for the reasoning.

Visitors who sign in through Purple Guest WiFi make a conscious-choice opt-in to marketing. Your contact list is therefore first-party and consented.

Purple runs across 80,000+ live venues and handled 440 million logins in 2024 (Purple data). The platform is ISO 27001 and Cyber Essentials certified, with 99.999% uptime. For sector context, see Purple for Retail.

The limit is real: WiFi will not give you an audited headcount at the door. Pair it with counters at your main entrances, and use those counts to calibrate.

When do mobile data panels earn their place?

Panels draw on two sources:

• aggregated mobile network data
• location data from apps whose members agreed to share it

They answer catchment questions no on-site system can: home postcode districts, travel time and visits to competing schemes. They need no hardware on site.

They are the wrong tool for daily operations, for three reasons:

• Counts are modelled from a sample, so small zones and quiet hours carry wide error.
• Reports usually lag the period they describe.
• Panels cannot see which mall or level a visitor used.

Use panels for acquisition due diligence, leasing pitches and annual catchment reviews, alongside your own counts.

What does each footfall method cost per entrance to run?

Vendors price per unit, per site or per report, so headline quotes rarely compare directly. Normalise every quote to a five-year total cost per entrance. Include installation, cabling, licences, maintenance and recalibration.

Method	Upfront cost per entrance	Recurring cost	Cost buyers often miss
WiFi presence analytics	No per-entrance hardware where access points exist	Software subscription	Calibration against a door count
Stereo cameras	Unit, mounting, power and data cabling to each door	Licence or maintenance contract per unit	Re-aiming after refits and shopfront changes
Thermal sensors	Unit, mounting and cabling per door	Licence or maintenance contract per unit	Checks on doors facing sun or heaters
Lidar sensors	Unit, high mounting and cabling per zone	Licence per sensor	Survey for suitable mounting points
Mobile data panels	None	Subscription per site or per report	Lag between visit and data delivery

Wide entrances change the arithmetic. A door that needs three stereo units costs roughly three times as much as a single-unit door. WiFi cost stays flat, because it scales with the venue rather than the doorway.

Worked scenario: extending a door count into zones (retail)

The figures below are illustrative.

Situation. An enclosed regional shopping centre has 140 units and 11 public entrances. Stereo counters cover seven mall entrances, and four car park lift lobbies have no counter. The asset manager needs dwell and visit share for the upper mall ahead of lease renewals. WiFi already covers every mall and the car park.

Approach. The team kept the seven counters and switched on WiFi presence analytics on the existing access points. Both systems ran side by side at the counted doors for six weeks. The team calculated the ratio of detected devices to counted entries at each door. They applied the average ratio to the four lift lobbies, then reported dwell and visit share by zone.

Outcome. The centre gained an estimated count for all 11 entrances without four new counter installations. Zone reporting covered all 140 units, grouped by mall and level. The monthly upper mall report showed visit share and median dwell, and fed the lease renewal pack. The team rechecks the ratio every quarter and after any network change.

How do you benchmark shopping centre footfall against the high street?

Footfall benchmarking in the UK typically draws on two published sources:

• BRC-Sensormatic IQ Footfall Monitor. Reports monthly year-on-year change for high streets, retail parks and shopping centres.
• MRI Software footfall indices. Cover UK retail destinations.

Both report percentage change, which tells you how to compare. Follow five rules:

1. Compare percentage change, not absolute volumes. Your 20,000 weekly entries mean nothing against a national index. Your year-on-year change does.
2. Use like-for-like entrances. If you add a counter mid-year, exclude it until it has 12 months of history.
3. Align calendars. Match trading weeks, Easter, bank holidays and school holidays. A moving Easter can swing a single month by several points.
4. Never mix methods. Compare door counts with door-count benchmarks. Keep WiFi device counts for internal trends, dwell and zone share.
5. Annotate local events. Roadworks, weather, a new competing scheme and anchor store closures all explain gaps that the index cannot.

Worked scenario: proving performance for a regeneration bid (public sector)

The figures below are illustrative.

Situation. A district council owns an open-air shopping precinct with nine street entrances and no doors. It is preparing a bid for regeneration funding. It needs evidence of how the precinct performs against national trends. Public WiFi already covers the precinct.

Approach. Doorway sensors were impractical, so the council installed lidar at the three busiest pedestrian gateways. WiFi presence analytics covered the rest of the precinct, calibrated against the lidar counts. From month 13, the team calculated year-on-year change from the three lidar gateways only. They set that figure against the BRC-Sensormatic IQ high street and shopping centre series.

Outcome. The council gained a monthly, like-for-like series aligned to the national benchmark, plus dwell by street from WiFi. Suppose the precinct rose 3% year on year while the high street series fell 1%. That four-point gap is evidence a funding panel can read at a glance. Dwell by street showed which frontages needed investment first.

How to decide for your estate

Match your situation to a starting point, then fill gaps over time.

Your situation	Recommendation
No door count, enclosed centre with defined entrances	Stereo cameras at every public entrance first
Door counters installed, WiFi across the malls	Add WiFi presence analytics for zones, dwell and return visits
Open-air retail park or precinct with no doors	Lidar at the main pedestrian gateways, plus WiFi across the scheme
Entrances in low light or with high privacy sensitivity	Thermal sensors at those entrances
Acquisition, valuation or leasing pitch	Mobile data panel for catchment, alongside your own door count
Portfolio of several centres	One door-count method and one WiFi platform across every site

If you run several centres, standardise early. Different counting methods at different sites make portfolio reports impossible to compare. The same central-control principle applies as in multi-site guest WiFi for franchises.

Questions to ask every footfall vendor

• What is your audited accuracy figure, how was it measured, and over what period?
• How do you validate counts after installation, and how often?
• What happens to the data when a sensor or access point goes offline?
• Is any image or identifier stored, where, and for how long?
• Who is the data controller, and will you sign a data processing agreement?
• Can I export counts by 15-minute interval through an API?
• Can I share tenant-level reports without exposing personal data?

Extra questions for WiFi vendors

• Which access point vendors do you support today, by name?
• How do you handle MAC randomisation and deduplicate devices?
• How do you keep marketing consent separate from network access?

Extra questions for camera and sensor vendors

• Does counting happen on the device, with no images leaving it?
• What door width does one unit cover at my mounting height?

Extra questions for panel vendors

• How many panel members fall in my catchment, and how did they consent?
• How do you model panel observations into total visits?

Frequently asked questions

Is WiFi footfall data accurate enough to replace door counters?

No, WiFi presence data should complement door counters at your main entrances rather than replace them. WiFi counts devices, not people, and phones randomise the address they broadcast, so absolute headcounts carry error. It is accurate for trends, dwell and zone share. Calibrate it against a door count over four to eight weeks. Then use the ratio to estimate uncounted doors and to measure the zones counters never reach.

Do we need new hardware to run WiFi analytics in our shopping centre?

No, not if your malls already have WiFi coverage. Purple runs as a cloud overlay on Cisco Meraki, HPE Aruba, Ruckus, Juniper Mist, Ubiquiti UniFi, Cambium, Extreme and Fortinet access points. You add software, not sensors. Coverage gaps, such as car park levels or service corridors, need extra access points only if you want data from those areas.

Is WiFi presence analytics compliant with UK GDPR?

Yes, when you run it with a lawful basis, a Data Protection Impact Assessment and clear signage. The ICO treats device identifiers such as MAC addresses as personal data where they can single out an individual. Hashing does not make them anonymous. Report in aggregate, set retention limits and keep marketing consent separate from access. Purple is ISO 27001 certified and GDPR compliant, with conscious-choice opt-ins for marketing.

Can we share footfall data with tenants?

Yes, share aggregated counts and zone-level dwell, never device-level records. Tenants value entrance trends and the footfall passing their unit. One access point often covers several units, so WiFi reports work best by zone or mall. Write the sharing terms into the lease or a data sharing agreement. Confirm that your vendor contract allows you to pass reports to third parties.

How does the cost of entrance counters compare with WiFi analytics?

Entrance counters cost more per entrance, because every doorway needs its own unit, mounting, cabling and maintenance. WiFi presence analytics on an existing network adds no per-entrance hardware, only a software subscription. Wide entrances push sensor costs up further, as one doorway may need several units. Ask every vendor for a five-year total cost per entrance, including recalibration after refits.

How do we benchmark our centre against the high street?

Compare percentage change, not absolute counts. The BRC-Sensormatic IQ Footfall Monitor publishes monthly year-on-year change for high streets, retail parks and shopping centres. Measure your own year-on-year change from the same entrances, using the same counting method in both years. Align trading weeks, Easter and bank holidays. Never benchmark WiFi device counts against sensor-based national figures.

Can we keep our existing camera counters and add WiFi?

Yes, and that combination gives you the most complete picture. Keep the counters as your headline entrance figure for reporting and benchmarking. Add WiFi presence analytics for dwell, zone flow and return visits, and use the counted doors to calibrate the WiFi ratio. Export both datasets by 15-minute interval into one reporting tool, so your centre and marketing teams work from the same numbers.