Shopping centre WiFi analytics platforms: footfall attribution and tenant reporting compared

Choose a platform that measures shoppers through authenticated WiFi logins, not passive MAC probes. Since iOS 14 and Android 10 randomise device addresses by default, probe counts overstate visitors and break repeat-visit data. Authenticated sessions give deterministic dwell, cross-shopping and return rates per mall zone. Entrance counters supply total footfall, and a capture-rate ratio joins the two.

Which shopping centre WiFi analytics platform should you choose?

For most regional centres, choose a hybrid: authenticated guest WiFi for behaviour, entrance counters for volume. Any shopping centre WiFi analytics platform comparison should start with one question. How does each option recognise a returning shopper?

Passive presence analytics counts device addresses. Authenticated analytics counts people who logged in. Since phones began randomising their addresses, that difference decides whether your repeat-visit and cross-shopping figures mean anything.

Three questions separate the platforms on a shortlist:

• Does it identify a visit by login identity or by MAC address?
• Can it map access points to mall zones, so dwell and cross-shopping report by tenant frontage?
• Can its visit data sit alongside your people-counter feed, so one capture rate links the two?

Purple's position is authenticated, hardware-agnostic and delivered as a cloud overlay on the access points you already run. Guest WiFi captures a conscious-choice opt-in at sign-in. WiFi Analytics then reports visits, dwell and returns against that identity. Purple runs across 80,000+ live venues in 90 countries and handles 500 million users and logins a year.

Where do footfall attribution methods genuinely differ?

Four approaches dominate retail destination WiFi reporting and the counting methods around it. They differ on what they count, how they handle randomised addresses and what you have to install.

Axis	Passive WiFi presence (probe counting)	Authenticated guest WiFi analytics	Camera or sensor counters	Hybrid: counters plus authenticated WiFi
What it counts	Device addresses seen by access points	Logged-in visitors and their sessions	People crossing a counting line	Gross entries plus a logged-in behavioural sample
Total footfall	Overstated where addresses rotate	Sample only, not a total	Accurate at counted entrances	Counter total, with behaviour scaled from WiFi
Dwell time	Unreliable once an address changes mid-visit	Session start to last access point seen	Not available: entry and exit are not linked	Session-based, reported against counted entries
Cross-shopping by zone	Fragmented by address changes	Zone-to-zone path per identity	Entrance only	Zone-to-zone path per identity
Repeat visits	Broken by per-network and rotating addresses	Linked to login identity across visits	Not available	Linked to login identity across visits
Effect of MAC randomisation	Severe	Low: identity anchors the visit	None	Low
Hardware needed	Existing access points	Existing access points	Sensor at every entrance, plus cabling	Existing access points plus entrance sensors
GDPR position	Device addresses are personal data, with no sign-in to give notice	Notice and consent captured at sign-in	Least personal data in the output when processed on the device	Consent at sign-in, count-only output from sensors
Who it suits	Continuity baseline during migration only	Centres measuring behaviour and loyalty	Centres reporting gross entries to investors	Asset managers reporting to tenants and investors

Why MAC randomisation breaks passive counting

A MAC (media access control) address identifies a device's network interface. Apple has given devices a private, per-network address by default since iOS 14. Android has randomised addresses per network by default since Android 10. Both also randomise the address in probe requests, the frames a phone sends while searching for networks. Recent releases of both can rotate the address on the same network too.

Passive presence tools count those addresses. One shopper can appear as several devices in a day, and as a new device on every visit. Visitor totals rise while repeat-visit rates fall. The dashboard still looks precise, but the numbers describe devices, not people.

How authenticated sessions measure dwell time

A login ties the session to an identity, not an address. A shopper connects through a captive portal, the sign-in page shown before internet access. The platform records who signed in and when the session began.

From there, RADIUS takes over. RADIUS is the protocol access points use to authenticate and account for each connection. Its accounting records, defined in IETF RFC 2866, report session start, interim updates and stop. Dwell runs from first access point seen to last, with a gap rule you set to close the visit.

If the address changes mid-visit, a fresh sign-in links the new session back to the same profile. Credential-based access goes further. IEEE 802.1X is the standard for port-based network access control, where the device presents a credential rather than relying on its address. Passpoint, built on the Hotspot 2.0 specification, lets devices join automatically with stored credentials. OpenRoaming, run by the Wireless Broadband Alliance, federates Passpoint across venues. With any of these, randomisation does not break the link between visit and identity.

How authenticated sessions measure cross-shopping

Map each access point to a zone: mall, food court, anchor entrance, car park lobby. A session moving between access points produces a sequence of zones. Aggregate those sequences into a zone-by-zone matrix. The matrix answers one question: of visitors seen in zone A, what share also reached zone B on the same visit?

Be precise with tenants about resolution. The centre's network usually covers common areas, not the inside of tenant units. Cross-shopping therefore reports by frontage zone, not by till transaction. Access point placement sets how finely you can split zones, so survey signal overlap between neighbouring zones before you name them.

How authenticated sessions measure repeat visits

Repeat visits are distinct visit days per identity over rolling windows, such as 30, 90 and 365 days. Because identity persists, a shopper who signed in last spring counts as returning today. That gives you frequency and recency cohorts, including lapsed visitors worth re-engaging through first-party data. Devices with saved credentials reconnect automatically, so returns register without another sign-in.

What reporting views do asset managers need for tenant leasing?

Asset managers report to three audiences. Investors want gross footfall and trend. Leasing teams want zone performance and anchor dependency. Tenants want evidence their frontage draws visitors. A shopping mall guest WiFi platform should serve all three from one data set.

Report view	Question it answers	Data source	Main audience
Centre footfall trend	Are entries up on last year?	Entrance counters	Investors, asset manager
Entrance split	Which car park or transit entrance feeds the centre?	Counters plus first zone seen in each session	Operations, leasing
Zone dwell	Where do shoppers spend time?	Authenticated sessions	Leasing, marketing
Cross-shopping matrix	Which zones share visitors on the same visit?	Authenticated sessions	Leasing
Anchor halo	What share of visits touch the anchor zone, then other zones?	Authenticated sessions scaled by capture rate	Leasing, investors
Repeat-visit cohorts	How often do shoppers return, and who has lapsed?	Login identity over time	Marketing, asset manager
Catchment	Which postcodes do visitors come from?	Postcode given at sign-in, with consent	Leasing, marketing
Event uplift	Did the event add visits and dwell?	Counters plus sessions against a baseline period	Marketing, sponsors

Anchor-store attribution

Define the anchor zone as the access points at the anchor's mall entrance. Measure the share of visits that include it. Then compare onward behaviour: how often do anchor visits continue to each other zone, against visits that never touch the anchor? The gap is an indicative halo effect.

Treat it as correlation, not proof. Shoppers who visit the anchor may differ from those who do not. Where you can, compare periods before and after an anchor change. Where leases require turnover reporting, set tenant sales alongside the zone data for a fuller picture.

Tenant reporting packs

Send tenants monthly, zone-level, aggregated packs. A useful pack shows frontage-zone visits, dwell against the centre average, the top cross-shopped zones and trend. Never share individual profiles. Suppress any figure built from a small group, so no individual can be identified, in line with the ICO's anonymisation guidance. Use the packs as context in renewals and rent reviews, alongside sales data rather than instead of it.

How do you combine WiFi login data with camera or sensor counters?

Login data is deterministic: each session belongs to a known identity. Counter data is anonymous and gross. Passive probe data is probabilistic at best. Combine the first two, and treat the third as a legacy series.

1. Validate the counters. Audit each entrance against manual counts at quiet and busy times. Counters can miss groups or double count in crowds, so record accuracy per entrance.
2. Align time and space. Use hourly buckets in one time zone. Map each counted entrance to its nearest access point zone.
3. Calculate the capture rate. Divide unique authenticated visitors by counted entries for each hour. Counters count entries, so a shopper who pops out to the car park counts twice. Use session data to estimate how often visitors leave and re-enter, then adjust.
4. Scale shares, not raw counts. Take behavioural shares from the WiFi sample, such as the share of visits reaching the food court. Multiply by counted entries to estimate totals.
5. Correct for sample bias. Shoppers planning a long visit are more likely to sign in. Weight by daypart and entrance, and report dwell as a sample metric rather than scaling it.
6. Monitor drift. The capture rate moves when you redesign the login page, add OpenRoaming or improve coverage. Annotate those changes so trends stay honest.

The blending itself happens in one of two places. Either the analytics platform ingests the counter feed, or both feeds land in the data warehouse your asset managers already use. Purple integrates with CRM and marketing platforms, so login data can reach the systems your marketing team runs.

Where each platform type fits

• Passive presence analytics fits only as a continuity baseline while you migrate. Do not use it for repeat visits or cross-shopping.
• Authenticated WiFi analytics alone fits smaller centres, outlet villages and retail parks with mall-wide coverage, where nobody contracts you to report gross entries.
• Counters alone fit centres that report gross entries to investors and need nothing else, or open-air schemes without WiFi in the walkways.
• The hybrid fits regional and super-regional centres reporting to both tenants and investors. See how Purple supports destinations on the Retail page.

All figures in the two scenarios below are illustrative, built to show the method rather than to report a named venue's results.

Worked scenario: an anchor exit at a regional centre

Situation. A 120-unit regional centre learns its department store anchor is leaving. Counters at six entrances recorded 8.4 million entries in the last 12 months. The leasing team needs to know how much traffic depends on the anchor before pricing the re-let.

What was done. The team mapped its existing access points into 16 zones. It took 12 months of authenticated sessions, with a capture rate averaging 18% of entries. It then ran the anchor halo and cross-shopping views.

Outcome. 41% of authenticated visits passed through the anchor zone. Of those, 57% continued into the east mall, against 29% of visits that never touched the anchor. Scaled to counted entries, roughly 3.4 million entries a year reached the anchor zone (0.41 x 8.4 million). The leasing team entered talks with a quantified anchor dependency and a named list of exposed east mall tenants.

Worked scenario: a six-week winter event in the atrium

Situation. A centre runs a six-week winter event in its central atrium. The sponsor and neighbouring tenants want evidence it added value.

What was done. The team compared counter entries with the same six weeks a year earlier. It mapped the atrium access points as an event zone. It then compared event-zone sessions against the centre-wide average for the same period.

Outcome. Entries rose 7% year on year. Visits that included the atrium averaged 84 minutes of dwell, against 61 minutes centre-wide. 23% of event-zone visitors were first-time sign-ins, and 48% went on to the food court. The report flagged calendar and weather differences, and the sponsor renewed on the strength of the dwell and new-visitor figures.

What does each option cost to run?

We cannot give a single price, because hardware reuse decides most of the bill. These are the cost drivers to model.

Option	Upfront hardware	Installation	Ongoing cost driver	Hidden cost
Passive presence	None beyond existing access points	Configuration only	Platform licence	Leasing decisions made on overstated counts
Authenticated WiFi analytics	None beyond existing access points	Configuration plus login page design	Platform licence	Login page work to lift the capture rate
Camera or sensor counters	One sensor per entrance, more on wide doors	Cabling and mounting at height, often out of hours	Maintenance and recalibration	Re-surveying after door or layout changes
Hybrid	Entrance sensors only	Sensors plus WiFi configuration	Two data feeds	Analyst time to maintain the capture-rate model

The biggest saving is reuse. Purple runs on Cisco Meraki, HPE Aruba, Ruckus, Juniper Mist, Ubiquiti UniFi, Cambium, Extreme and Fortinet. The case for avoiding a hardware refresh is set out in Why your next WiFi upgrade doesn't need new hardware. If you run a controller-based estate, read What is a WLC (Wireless LAN Controller) and Do You Still Need One?. Ruckus estates should read Ruckus SmartZone and RUCKUS One guest WiFi vs external captive portal platforms.

Budget for coverage as well. Shoppers who cannot connect cannot sign in, and every dead spot lowers your capture rate.

How to decide for your estate

Match your reporting obligations to the option.

Your situation	Recommendation
You report gross entries to investors and nothing else	Entrance counters now; add authenticated WiFi when leasing needs behaviour
You have WiFi across the malls but no counters	Authenticated WiFi analytics now; counters at main entrances when you need totals
You run passive presence analytics today	Move to authenticated sessions and run both for one full trading quarter
You face an anchor exit or major void	Hybrid, with anchor halo and cross-shopping views
You run a portfolio of centres	Hybrid on one hardware-agnostic platform, so reports compare like for like

Then put these questions to every vendor on your shortlist:

• Is a visit keyed to login identity, a credential or a MAC address?
• Can you define zones by access point and rename them as tenants change?
• Which credential methods are supported: captive portal, IEEE 802.1X, Passpoint, OpenRoaming?
• How are consent and marketing opt-ins recorded, and can a shopper withdraw them?
• Which certifications cover the data? Purple holds ISO 27001, GDPR, CCPA, Cyber Essentials and B Corp certification.

Finally, follow the ICO's guidance on WiFi location analytics. Clear signage and a plain-language notice at sign-in protect your data, and they keep your capture rate high.

Frequently asked questions

Does WiFi analytics still work now that phones randomise MAC addresses?

Yes, provided the platform measures authenticated sessions rather than passive probes. A login ties each visit to an identity, so a new device address does not create a new visitor. Credentials such as IEEE 802.1X, Passpoint and OpenRoaming authenticate the device without relying on its address. Passive presence tools lose accuracy because iOS and Android randomise addresses by default, which overstates visitors and breaks repeat-visit counts.

Can Purple run on the access points our shopping centre already has?

Yes. Purple is hardware-agnostic and runs as a cloud overlay on Cisco Meraki, HPE Aruba, Ruckus, Juniper Mist, Ubiquiti UniFi, Cambium, Extreme and Fortinet access points. You point your existing network at Purple for guest sign-in and analytics, with no rip and replace. If coverage is weak in parts of the mall, fix that first: shoppers who cannot connect cannot sign in.

Do we still need people counters if we have guest WiFi analytics?

Yes, if you report gross footfall to investors or tenants. Authenticated WiFi measures the shoppers who sign in, which is a sample of total entries. Counters at each entrance measure the total. A capture rate joins the two, so behaviour seen in the WiFi sample scales to the counted population. Without counters, present WiFi figures as sample-based metrics, not totals.

How is Purple licensed for a shopping centre?

Purple licenses through three plans: Connect, Capture and Engage. Because Purple runs on the access points you already own, you avoid a hardware refresh to start measuring. Entrance counters, if you add them, are a separate hardware cost. Ask Purple for a quote covering your estate, including the number of centres and the access point vendor at each one.

Is shopping centre WiFi analytics compliant with GDPR?

It can be, when you collect data transparently and with a lawful basis. Login identifiers and device addresses are personal data under GDPR. Purple captures conscious-choice opt-ins at sign-in, so marketing consent is a deliberate, separate choice. Purple holds ISO 27001, GDPR, CCPA and Cyber Essentials certification. Share only aggregated, zone-level reports with tenants, and follow the ICO's guidance on WiFi location analytics.

Can we share WiFi analytics reports with our tenants?

Yes, at aggregated zone level. Tenant packs typically show frontage-zone visits, dwell against the centre average, the top cross-shopped zones and month-on-month trend. Never share individual profiles or contact details, which stay with you as first-party data under your privacy notice. Suppress any figure built from a small group, so no individual can be identified, in line with ICO anonymisation guidance.

How do we migrate from a passive presence analytics tool?

Run both systems in parallel for one full trading quarter, then switch. Under MAC randomisation, passive tools overstate visitors and understate returns. Expect visitor counts to fall and repeat-visit rates to rise once you move to authenticated data. The parallel period lets you explain that step change to investors and tenants. Keep counter data as your continuous footfall series throughout.