There is no single best way to measure foot traffic. The right method depends on whether you need to count people at a single entrance, understand movement patterns inside a store, or benchmark visit trends across thousands of locations you have never set foot in.

Most guides list the methods and stop there. The more useful question is which method answers which business question, because the team evaluating a new site for a 10-year lease has completely different data needs from the store manager trying to optimise Saturday staffing.

This guide covers eight methods for measuring foot traffic, what each one actually captures, what it misses, and how modern retailers combine them.

Two categories of measurement

Before comparing individual methods, it helps to understand the fundamental split in how foot traffic can be measured.

Operational measurement tells you what is happening inside your own locations. Hardware installed at your entrances, on your ceiling, or in your WiFi network counts the people walking through your specific doors. The data is precise for that location, but it tells you nothing about competitors, the broader market, or where your visitors came from.

Strategic measurement tells you what is happening across the market. Third-party data platforms model visits to millions of locations using aggregated mobile device signals. The data is broader and enables competitive benchmarking, trade area analysis, and market-level trends, but it is modelled rather than directly observed.

These are not competing approaches. They answer different questions and serve different teams. Store operations needs operational data to run individual stores. Real estate, marketing, and strategy teams need strategic data to make portfolio-level decisions. Most mature retailers use both.

Method 1: Manual clickers

The simplest approach. Staff members stand at entrances with handheld tally counters and press a button for each person who enters.

What it captures: Raw headcount at a single entry point during the hours someone is counting.

What it misses: Everything else. No demographic data, no dwell time, no repeat visit tracking, no competitor insights. It also misses anyone entering when the counter is not staffed, and it is prone to human error, especially during busy periods or when staff are multitasking.

Best for: Temporary setups like pop-up shops, events, or one-off traffic checks where installing hardware is not justified. Some retailers use manual counts to validate automated systems during a calibration period.

Cost: Essentially free (the clickers cost a few pounds), but the labour cost of dedicating a person to counting adds up quickly.

Accuracy: Low. Studies consistently show manual counts diverge from automated counts by 10-20%, and the error compounds during peak periods when accuracy matters most.

Method 2: Infrared beam sensors

A pair of devices installed on either side of a doorway creates an invisible infrared beam. When someone walks through and breaks the beam, the system registers a count.

What it captures: Directional entry and exit counts. Better systems use dual beams to determine direction (in vs out) and can distinguish between two people entering side by side.

What it misses: Anything beyond a headcount. No demographic data, no in-store movement, no dwell time. Single-beam systems struggle with groups entering together and can double-count people who pause in the doorway.

Best for: Retailers who need a reliable, low-maintenance entrance counter for individual stores. The hardware is relatively affordable and lasts years with minimal upkeep.

Cost: Roughly $200-$800 per doorway, plus installation and any software subscription for data aggregation.

Accuracy: 85-95% depending on the system and doorway configuration. Accuracy drops in wide entrances and during peak traffic when visitors enter in clusters.

Method 3: Thermal sensors

Thermal counters detect body heat signatures as people pass through a zone. Unlike beam sensors, they can count from above (ceiling-mounted) and do not require hardware on both sides of an entrance.

What it captures: Headcount with reasonable accuracy in both directions. Ceiling-mounted thermal sensors can also track basic movement paths within a zone, not just entry/exit.

What it misses: Demographics, visit duration beyond the sensor zone, repeat visitor identification. Performance degrades in very hot environments where ambient temperature approaches body temperature.

Best for: Retailers who want more flexible installation than beam sensors (especially for wide or unusual entrances) and are willing to pay a modest premium for better accuracy.

Cost: $300-$1,500 per unit depending on the system. Solutions like Dor offer thermal sensors with POS integration.

Accuracy: 90-97%. Generally more accurate than beam sensors because they are less affected by groups and doorway geometry.

Method 4: Video analytics and AI cameras

Camera systems (often repurposing existing security cameras) combined with computer vision software to count people, track movement patterns, and in some cases estimate demographics like age range or group size.

What it captures: The richest single-location dataset of any hardware method. Headcounts, movement paths, dwell time by zone, queue lengths, heatmaps showing high-traffic areas within the store. Advanced systems can track the customer journey from entrance to specific departments.

What it misses: Anything outside the camera’s field of view. No competitor data, no trade area information, no visitor origin. Privacy regulations limit what demographic data can be collected, and processing video at scale requires significant compute infrastructure.

Best for: Large-format retailers, shopping centres, and airports where understanding in-store movement patterns justifies the investment. Also valuable for layout optimisation and queue management.

Cost: $500-$2,000+ per month depending on the number of cameras and the analytics platform. If you are leveraging existing security cameras, the marginal cost is the software subscription.

Accuracy: 95-99% for headcounts in well-configured environments. Lower for more complex metrics like dwell time and path analysis.

Method 5: WiFi and Bluetooth tracking

When visitors carry smartphones with WiFi or Bluetooth enabled, nearby access points can detect their devices. This allows retailers to track presence, movement, and repeat visits within a location.

What it captures: Device presence, approximate dwell time, repeat visit frequency, and movement between zones (if multiple access points are deployed). Can leverage existing WiFi infrastructure, reducing setup costs.

What it misses: Not everyone has WiFi or Bluetooth enabled. Modern smartphones use MAC address randomisation, which makes it harder to track unique devices accurately. Devices can be detected outside the store (in a car park or adjacent unit), leading to overcounting. No demographic data without additional integrations.

Best for: Shopping centres and large stores that already have WiFi infrastructure and want zone-level movement data without installing additional hardware. Also used for captive portal marketing (where visitors log in to WiFi and provide an email).

Cost: $100-$500 per month for analytics software if leveraging existing WiFi. More if deploying dedicated Bluetooth beacons.

Accuracy: 60-80%. The least precise of the automated methods due to MAC randomisation, signal variability, and the gap between “device detected” and “person actually in the store.” Accuracy has declined over the past several years as Apple and Android have tightened privacy controls.

Method 6: Pressure mats and floor sensors

Mats placed at entrances that detect footsteps through weight or pressure changes.

What it captures: Basic entry/exit counts.

What it misses: Nearly everything. No direction, no demographics, no in-store movement, no dwell time. Can be triggered by trolleys, pushchairs, and staff.

Best for: Very low-budget deployments where any data is better than none. Increasingly rare in professional retail settings.

Cost: $50-$300 per mat.

Accuracy: 70-85%. The lowest accuracy of any automated method.

Method 7: POS and transaction data as a proxy

Using point-of-sale records to infer visit volume. If you sold 200 transactions today, at least 200 people visited. Combined with a known conversion rate, you can estimate total traffic.

What it captures: A floor for visit volume (actual visits will always be higher than transactions). When combined with a conversion rate derived from another counting method, it provides a cross-check on traffic estimates.

What it misses: Every visitor who did not buy. Browsers, returners, comparison shoppers, and people who left because of a queue are all invisible. It also conflates single-item purchases with multi-item baskets, and cannot account for multiple visits by the same person.

Best for: Retailers who have no counting hardware but want a rough directional indicator. Also useful as a validation layer alongside other methods.

Cost: Zero incremental cost if you already have a POS system.

Accuracy: As a standalone traffic estimate, very low. As a cross-check against sensor or mobile data, it is useful for validating conversion rate calculations.

Method 8: Mobile location data platforms

Third-party platforms (including PassBy, Placer.ai, and others) use aggregated, anonymised signals from mobile devices to model visit volumes across millions of locations. GPS, WiFi positioning, Bluetooth, and cellular signals from a panel of devices are cleaned, modelled, and scaled to represent the full population.

What it captures: This is the only method that provides both depth and breadth. Visit counts, unique visitors, dwell time, visit frequency, trade area (where visitors come from), demographic and psychographic profiles, competitive benchmarking, and predictive forecasts, all without installing any hardware.

What it misses: It is a modelled estimate, not a direct observation. Accuracy depends on the panel size, the quality of the POI database (how well store boundaries are defined), and the sophistication of the modelling. Visitors without smartphones or with location services disabled are not captured by the panel, though statistical modelling adjusts for this.

Best for: Any team that needs to see beyond their own four walls. Real estate teams evaluating sites, marketing teams measuring campaign impact, strategy teams benchmarking against competitors, and finance teams building alternative data models all rely on mobile location data because no other method provides cross-market, cross-competitor visibility.

Cost: Varies by provider and scope. Platform access typically starts in the thousands per year. API and data feed access for enterprise deployments is priced based on the number of locations and data depth required. PassBy offers Essential, Premium, and Ultimate plans. See pricing →

Accuracy: Varies significantly by provider. PassBy achieves 94% correlation to ground truth by validating against in-store sensors and sales data across hundreds of thousands of locations. This validation step is critical. Without it, a mobile data platform is just a black box making estimates with no accountability.

How to choose the right method

The decision tree is straightforward:

If you need to count people at a single entrance for staffing and operations: thermal sensors or video analytics, depending on budget and whether you need in-store movement data.

If you need to understand how people move inside a single location: video analytics or WiFi tracking, depending on whether you need zone-level path data or just dwell time.

If you need to compare locations, benchmark competitors, analyse trade areas, or evaluate sites you have never visited: a mobile location data platform. No hardware method can do this.

If you need all of the above: use a mobile data platform for strategic measurement and layer in hardware at key locations for operational granularity. The mobile data provides the market context. The hardware provides the in-store detail. Together they give you a complete picture.

For a comparison of the major mobile data providers, see our guide to foot traffic data providers. For a deeper look at what mobile location data includes and how to evaluate quality, see our foot traffic data guide.

Explore how PassBy measures foot traffic across millions of US retail locations. Book a 15-minute walkthrough →

FAQ

What is the most accurate way to measure foot traffic? For a single entrance, ceiling-mounted thermal sensors or AI-powered video analytics offer the highest accuracy (95-99%). For cross-market analysis covering thousands of locations, mobile location data platforms validated against ground truth are the most reliable. PassBy achieves 94% correlation to ground truth across its coverage.

How do you measure foot traffic without sensors? Mobile location data platforms like PassBy measure foot traffic using aggregated signals from mobile devices, without any hardware installation. This is the only method that works for locations you do not own or have not visited, which makes it essential for competitive analysis and site selection.

What is a foot traffic counter? A foot traffic counter is any device or system that counts the number of people entering a location. This includes manual clickers, infrared beam sensors, thermal sensors, video cameras with analytics software, WiFi detection systems, and pressure mats. Each varies in accuracy, cost, and the depth of data it provides.

How much does it cost to measure foot traffic? Costs range from nearly free (manual clickers, POS proxy) to hundreds per month (sensors, WiFi analytics) to thousands per year (mobile data platforms). The right investment depends on what questions you need to answer. Counting heads at one door is cheap. Understanding competitive dynamics across a national portfolio requires a more comprehensive data source.

Can you measure foot traffic for a competitor’s store? Not with hardware methods, which require physical installation at the location. Mobile location data platforms are the only way to measure competitor foot traffic, because they model visits from aggregated device signals without needing any access to the competitor’s premises.

How do retailers combine multiple measurement methods? The most common combination is mobile data for strategic insight (site selection, competitive benchmarking, trade area analysis) and in-store sensors for operational detail (hourly staffing, layout optimisation, queue management). The mobile data tells you how a store performs relative to the market. The sensors tell you what is happening inside that specific store. Some retailers also use their POS data as a validation layer, comparing transaction-derived conversion rates against sensor counts to ensure consistency.