Expert guide · People counting
People counters explained: the 2026 guide to foot traffic counting
What a people counter is, how the sensors actually work, what an accuracy figure does and does not promise, what a system costs in 2026, and where the privacy line sits under GDPR, the EU AI Act and North American law. Written to answer the questions people actually ask — with the sources linked.
Published by SMS Storetraffic. We have made and sold people counters since 1972, so we have a stake in you buying one. Where a claim is ours, it is labelled; where it is someone else’s, it is linked. Read accordingly.
A beam sensor, a 3D camera and the software that turns their counts into decisions — our platform
The short answer: a people counter is a sensor mounted at an entrance that records how many people pass, usually in each direction, and sends the totals to software. In 2026 the sensors are infrared beams, thermal arrays, 3D stereo cameras and time-of-flight depth sensors; good ones count 95–99% of people, cost $300–$2,000 per door plus $0–$150 a month for software, and — when they are designed to count rather than to watch — process no personal data at all.
What it is forTurning sales into a rate. Traffic × conversion rate × average sale = sales. Without the traffic number, a slow week and a badly staffed week look identical on the till.
What accuracy meansA ceiling (“up to 99%”) and a floor (“never below 95%”) are opposite promises. Ask for the floor, and for how it is verified. Consistency matters more than the last percentage point.
What it costsWireless beam sensors from about $300–$500; 3D sensors from about $300, most $650–$1,200, outdoor and enterprise units to $2,000; software from $0 to about $150 per location per month. The cable run to the door, when one is needed, is the cost nobody quotes.
Is it legalYes. A counter that stores no images and identifies nobody processes no personal data under GDPR, is untouched by the EU AI Act’s prohibitions, and collects no biometric identifier under Illinois BIPA or Quebec’s Law 25. Camera systems that keep video are a different matter.
Manual countingUndercounts by 8–25% in the one peer-reviewed comparison we know of, and costs more in wages within a week than most sensors cost outright.
Definition
What is a people counter?
Definition
A people counter is an electronic sensor installed at a doorway or passage that detects each person crossing an invisible line and records the crossing, usually with its direction. The sensor is the hardware half of a people counting system; the software half stores the counts by time of day, combines them with sales, staffing and weather, and reports them.
The same device goes by many names depending on country and industry: footfall counter (United Kingdom, Europe), foot traffic counter or store traffic counter (North America), door counter, visitor counter, customer counter, patron counter (libraries), attendance counter (worship, events) and, in French, compteur de personnes or compteur d’achalandage. The metric they all produce is traffic: the number of people who entered a defined space in a defined period.
Two things a people counter is not. It is not a security camera: it does not record who came in, and a purpose-built counter stores no images at all (more on that under privacy). And it is not a mobile-location dataset: services that estimate visits from smartphone location data describe a market; a counter measures your door.
People counters started in retail — the first United Kingdom beam counters were installed at Meadowhall in 1988, covering 14 entrances (Wikipedia) — and retail is still the largest user. But the same sensors now count library patrons for annual reporting, museum visitors for grant applications, congregations, gym members, office occupancy and passengers.
Mechanism
How do people counters work?
Every automated people counter does the same three things: it detects a person crossing a count line, it decides whether that crossing is a valid count (and in which direction), and it records the count with a timestamp for software to aggregate. What differs between technologies is how the detection is done — and therefore what confuses it.
Detect
A horizontal infrared beam is broken; a thermal array sees a warm shape move across a cold floor; a stereo camera computes the height of everything below it and sees a head-shaped object at 1.6 m; a time-of-flight sensor measures the distance to the floor change. Each is a different way of answering “is a person here?”
Decide
The sensor’s firmware turns detections into counts. This is where the intelligence lives: was that one person or two side by side? A child, a cart, a stroller? Someone who walked in and straight back out? A person standing in the doorway for a minute? Basic sensors apply a rule (anything that breaks the beam is one count); 3D and AI sensors track each object across the field of view and count it once, when it crosses the line, in the direction it went.
Record and report
Counts are timestamped, buffered on the device and sent to software — by Wi-Fi, Ethernet or cellular — at intervals from real time to hourly. The software does the useful work: hourly and daily totals, business-hours filtering, comparison to sales and staff hours, alerts when a sensor goes quiet.
The five sensing technologies in use in 2026 are summarised below. Each has a full section, with what breaks it and where it belongs, in the technology comparison.
| Technology | How it detects | Counts groups side by side | Filters children / carts | Typical accuracy claim | Power | Privacy |
|---|---|---|---|---|---|---|
| Horizontal infrared beam | A transmitter and receiver across the doorway; a broken beam is a count | No — two abreast is one count | By mounting height only | 95–98% on single-file adult traffic | Batteries or wired | Sees nothing; no image exists |
| Overhead thermal | A low-resolution heat map; warm shapes tracked across a cold floor | Yes, until shapes overlap | Poorly — heat, not height | 95–99% claimed | Wired or battery + cellular | No recognisable image |
| 3D stereo vision | Two lenses compute a height map; heads are tracked as objects | Yes | Yes — by height and shape | 95–99%, some with a guaranteed floor | PoE (one cable) | Depth data; video off by default on good systems |
| Time-of-flight / LiDAR | Infrared pulses timed to the floor; a depth map with no colour image | Yes | Yes — by height | 95–98% claimed | PoE | Geometry only; no image possible |
| Wi-Fi / Bluetooth sniffing | Counts phones broadcasting nearby | Counts devices, not people | No | Estimates only since MAC randomisation | Wired | Device identifiers — regulated |
Accuracy figures are vendors’ published claims for their own products under good conditions; see what accuracy means. “Privacy” describes what the sensor captures by design, not how a given installation is configured.
Purpose
Why count people at all?
Because sales figures tell you what happened, and only traffic tells you why. A store that sold less this week either had fewer people walk in, or converted fewer of the people who did, or sold them less each. Those are three different problems with three different fixes — marketing, staffing and service, merchandising — and a till receipt cannot tell them apart. This is the Retail Equation:
Traffic × Conversion rate × Average sale = Total sales
Two of the three terms come from the point of sale. The third only exists if something counts the people who did not buy. That is the whole case for a people counter: it is the only instrument that measures the opportunities you did not convert.
The stakes have not moved online. E-commerce was 17.1% of U.S. retail sales in the second quarter of 2026 (U.S. Census Bureau), so roughly five dollars in six are still spent in a physical store. And the store visit itself is changing: ICSC and McKinsey’s April 2026 report, based on a survey of 3,004 U.S. consumers, finds shopping trips becoming less frequent and more intentional as people research before they leave home (Shopping in the Age of AI). Fewer, more deliberate visits make each one worth more — and make knowing how many you had, and how many you converted, matter more rather than less.
Your first conversion rate is a starting point, not a verdict. Almost everyone is disappointed by it. Most mall-based specialty stores sell to fewer than half the people who walk in; many sell to fewer than one in ten (our own experience since 1972). The number does not matter; the trend does. A store that moves from 10% to 11% conversion has grown sales 10% with the same rent, the same traffic and the same marketing spend.
12–20%
where most stores land the first time they count
Managers who believed they converted 80% of visitors discover it is 12–20% once a counter is installed (our archive). Four more sales per hundred visitors on a 20% base is a 20% sales increase.
2% → 7%
closing ratio over three years
An international apparel chain that installed counters to set staff-to-traffic ratios, then added hours at peak periods while competitors cut them. Sales rose and wage cost fell as a share of sales (our archive).
$1.45B
people-counting market, 2025
Growing about 13% a year to roughly $3B by 2031 (Mordor Intelligence; Fortune Business Insights puts 2025 at $1.46B). Retail is the largest segment; occupancy analytics the fastest-growing driver.
What is a good retail conversion rate?
There is no single answer, and be suspicious of anyone who gives one without a vertical. In our experience, store managers who start counting usually discover a conversion rate somewhere between 12% and 20% — far below the 50% or more most of them believed (our archive). Most mall-based specialty stores sell to fewer than half their visitors and many to fewer than one in ten. Grocery converts most visits because people arrive intending to buy; luxury converts least because people arrive intending to look. The independent ranges below are the ones Shopify publishes in its 2026 retail guide; no sample sizes are given, so treat them as directional.
| Store type | Typical conversion | Why |
|---|---|---|
| Grocery | 20–40% | High-intent visits: people arrive to buy a known list |
| Specialty retail | 15–30% | A mix of intent and browsing; the range where most of our customers start |
| Big-box formats | 10–20% | Large footprints, comparison shopping, many visits that end without a basket |
| First-time counters, all formats (our experience) | 12–20% | What managers usually find in the first month, against a belief of 50% or more |
Format ranges from Shopify, Retail Conversion Rate (updated March 2026); the first-time-counter range is our own observation, first published in 2016. For comparison, Shopify puts online conversion at roughly 1.6–3%.
Beyond retail: where a visitor count is the product
For a growing number of institutions the count itself is the deliverable. Public libraries in the United States report “library visits” every year through the IMLS Public Libraries Survey, which defines a visit as the total number of persons entering the library for whatever purpose — and IMLS has acknowledged that reported counts can be inaccurate because of limited counting technology and unrepresentative “typical week” sampling multiplied by 52 (our guide to the IMLS requirements). Museums use attendance to support grant applications. Places of worship track attendance trends; offices measure desk and meeting-room utilisation; gyms, transit and event venues manage live occupancy against a capacity limit. The sensors are the same; the software emphasis shifts from conversion to reporting and occupancy.
Decisions
What do you actually do with foot traffic data?
Staff to traffic, not to sales
Traffic by half-hour tells you when people are in the store, which is not when the till is busiest — the till is busiest after the queue has already formed. Scheduling one associate per ten visitors per half-hour (a common service target) from IN counts produces a materially different rota from scheduling on sales history, and it is the single most reliable lever on conversion. Our worked example shows why IN counts beat “people in store” for this.
Measure marketing by the door
A campaign either brought more people to the door or it did not. Comparing traffic in campaign weeks against the same weeks last year, by hour, is the only clean measure of local marketing, window displays, events and opening-hours changes — sales confound it with conversion and price.
Find the key selling periods
Every store has hours where conversion is double the average and hours where it halves. Those are usually staffing or fixture problems in disguise. Traffic by hour against transactions by hour finds them in a week; instinct never does.
Capture rate and layout
A second sensor facing the mall corridor or the street gives passers-by; entries divided by passers-by is your capture rate — the window display’s conversion rate. Inside, counters on department thresholds show which areas are visited and which are walked past.
Live occupancy
IN minus OUT, in real time, against a limit — for fire-code capacity, gyms, libraries and study spaces, event floors. Works well with 3D sensors on every entrance and a reset each morning; see the caveat about drift below.
Prove impact
For libraries, museums, community centres and places of worship, a continuous, timestamped count is the difference between “we estimate 40,000 visits” and a defensible annual figure for boards, funders and the Public Libraries Survey.
The honest section
How accurate are people counters?
Modern 3D and thermal counters, installed properly, count between 95% and 99% of the people who cross the line; beam counters count 95–98% of single-file adults and less when people walk abreast. Those numbers are less useful than they look, for three reasons.
1. A ceiling is not a floor. “Up to 99%” describes the best day the vendor has seen. “Never below 95%, verified by video, or we fix it” describes the worst day you will accept. Only a minority of vendors in our nine-system comparison commit to a floor at all; we back the 3D Scope II LC with a 95% floor for two years, verified by video. Most of the market says “up to”. Always ask for the minimum a vendor will stand behind and the method by which it is verified.
2. Consistency matters more than accuracy. A counter that catches 90% of people every day is a fine instrument: your Saturdays are still bigger than your Tuesdays, this year is still comparable to last, and every ratio built on it is stable. A counter that catches 100% one day and 70% the next is useless for decisions, whatever its average. This is why beams, which systematically miss the same kind of crossing, remain a legitimate entry-level choice, and why an unreliable network or a battery that keeps dying does more damage than a percentage point of accuracy.
3. Accuracy is measured against a definition. Are children counted? Staff? Someone who steps in, turns around and leaves? A person who stands in the doorway for two minutes? Two counters can both be “98% accurate” against different rules. Agree the rules before the audit.
What causes counting errors
- People abreast — a beam counts a couple as one; overhead sensors separate them by head.
- Lingering in the count zone — greeters, queues and displays at the threshold generate repeat or missed counts unless the sensor tracks objects.
- Children, carts, strollers, dogs — counted unless filtered by height (beam at 54 in / 137 cm; overhead sensors by measured height).
- Doors — a door swinging through a beam is a count; a sensor mounted over the swing sees a moving slab.
- Light — direct sun on a beam receiver blinds it (our PEARL drops from 15 ft to 7 ft of range in bright sun); hard shadows fool 2D video; below ~3 lux a stereo camera loses depth; thermal is indifferent to both.
- Heat — an open door on a cold day, a heater or a sunlit floor can confuse thermal and passive-infrared sensors.
- Standing still — older thermal sensors could lose a person who stood motionless for a few seconds and re-acquire them on the wrong side of the line.
- Staff — a busy team crossing the line 200 times a day is not customer traffic; exclude them with UWB tags or a separate staff door.
Why live occupancy drifts
Occupancy is IN minus OUT, and every error accumulates. A sensor that is 95% accurate at a door seeing 100 people an hour makes about five errors an hour. If those errors lean one way — and they do, because the causes are systematic — the building “contains” 25 more people than it does by the end of a 12-hour day, and 60 by the end of a weekend without a reset. That is arithmetic, not a defect. It is why occupancy systems need the most accurate sensor on every entrance, a daily reset, and a manual correction button; and why, for a shop that simply wants traffic, directional counting adds cost without adding much decision value (the staffing evidence).
How to audit a counter
Record 15–30 minutes of video at a peak period with the count lines overlaid, count the people on the video yourself (rewinding where you are unsure), and compare with the sensor’s total for the same window. A good sensor records this itself; the 3D Scope II LC stores an auditable recording on request and our auditing policy describes the process. Never audit a counter from a chair with a clicker — the person is the less reliable instrument (see manual vs automated).
People versus sensors
Manual counting vs automated counting
A person with a clicker is the least accurate and most expensive people counter you can buy. In the one peer-reviewed comparison we know of, manual counts with tally sheets or clickers systematically undercounted by 8–25% against video of the same crossings, with error worst at the start and end of each session (Diogenes, Greene-Roesel, Arnold & Ragland, Transportation Research Record 2002, 2007 — source).
The cost comparison is not close. At the U.S. median retail wage of $17.03 an hour (BLS, May 2025), a person at the door for a 62-hour trading week costs over $1,050 in wages alone. A wireless beam counter costs less than that once; a 3D sensor pays for itself in about a week. And the person still undercounts.
Manual counting keeps two legitimate jobs: a short video audit to validate a sensor, and a one-off count where there is no infrastructure at all. The full argument, with the seven reasons a machine wins and the audit protocol, is in Manual vs automated people counting
8–25%undercount by clicker or tally sheet vs video (peer-reviewed, 2007)
$1,056a week in wages for one door, 62 hours, U.S. median retail wage
<30 hof door-wages to cover a $499.95 wireless beam counter outright
Budget
How much does a people counter cost in 2026?
For one entrance, expect $300–$500 for a wireless beam sensor, $300–$1,200 for a 3D stereo or time-of-flight sensor (most fall between $650 and $1,200), up to about $2,000 for enterprise-grade or outdoor units, and $0–$150 per location per month for software. Three of the largest vendors publish no prices at all. Our nine-system comparison found hardware from $299 to $1,149.95 where it is sold outright, with quote-only enterprise units above that; consumer-grade beam counters with a display sell online for under $100, with the limitations described under tally counters.
| Cost line | Typical 2026 range | What decides it |
|---|---|---|
| Sensor hardware | $300–$500 beam · $300–$1,200 3D/ToF (most $650–$1,200) · $1,500–$2,000 outdoor or enterprise | Technology, weather rating, whether bought or bundled into a subscription |
| Software | $0 (free tiers) to about $150 per location per month | History depth, granularity, POS integration, real-time occupancy, number of users and reports |
| Getting power to the door | $0 for battery sensors · $150–$400 per PoE drop · $300–$850 for a quoted one-cable job | Whether the sensor needs a cable at all; ceiling type; after-hours contractor premium. Almost never in the vendor’s quoted install time (detail) |
| Installation and calibration | $0 self-install to several hundred dollars per site for professional installation | Peel-and-stick beams need nobody; 3D sensors need a ladder and a remote calibration; enterprise rollouts include a site survey |
| Support and warranty | Included, or an unpublished annual support contract | Read the terms: some vendors make portal access conditional on an active support contract |
| Shipping, duties, tax | $0 to $55 shipping; duties vary | Where the vendor ships from and whether they cover import duties for your country |
Think in five-year totals, not sticker prices. A $300 sensor on a $150-a-month plan costs about $8,400 over five years for one door; a $500 sensor on a free plan costs about $525 including batteries. Neither is wrong — the expensive one may include cellular connectivity, hardware replacement and analytics you need — but they are not the same purchase, and the cheaper day-one number is the dearer one to own. Every published figure for nine systems, over three and five years, is in the comparison.
Return on investment is a conversion question. Take your weekly traffic, your conversion rate and your average sale. Ask what one extra sale per hundred visitors is worth per year. For almost any store with more than a few hundred visitors a week, that single point of conversion pays for the counter, the software and the cable within months — provided someone acts on the data. A counter nobody looks at has an ROI of zero.
Compliance
Are people counters legal? Privacy under GDPR, the EU AI Act and North American law
Yes, people counters are legal everywhere we sell — and a counter built to count rather than to watch sits outside most privacy law altogether, because it never collects personal data. The legal question is not “is counting allowed” but “what does this particular sensor capture and keep”. Three categories:
Anonymous by design
Beams, thermal, time-of-flight, 3D depth
No image of a face can exist because none is captured: a beam sees a break, a thermal array sees a warm blob, a ToF sensor sees distances, a stereo counter reduces its frames to a height map on the device. What leaves the sensor is a number with a timestamp. This is not personal data under GDPR, not a biometric identifier under Illinois BIPA or Texas CUBI, and not “renseignements personnels” under Quebec’s Law 25.
Camera-based, configured well
Video and stereo sensors with video features
Many 3D counters contain real cameras and can record video for installation checks and audits. When that is off by default, stored briefly on the device and never used to identify anyone, regulators treat it as low risk; a short data-protection impact assessment is still good practice in the EU and UK. Our 3D Scope II privacy notes describe the defaults: no video or image transfer in normal operation, no face recognition, recording optional and removable.
Identifying
CCTV analytics, face-based demographics, Wi-Fi tracking
Re-using security cameras for counting brings the whole CCTV rulebook (signage, retention, access). Estimating age or gender from faces is biometric processing. Wi-Fi and Bluetooth sniffing collects device identifiers, which the EU and Quebec regulators have treated as personal data. All can be done lawfully; none is “just a counter”.
European Union and United Kingdom
Under GDPR, data is personal if it relates to an identifiable person. Anonymous counts do not: Recital 26 states that the principles of data protection “should not apply to anonymous information” and that the Regulation “does not concern the processing of such anonymous information, including for statistical purposes” (GDPR Recital 26). So the heaviest obligations (lawful basis, DPIA, subject-access rights) do not attach to a beam, thermal or depth-only counter. The EU AI Act’s prohibited practices have applied since 2 February 2025: they ban biometric categorisation by sensitive attributes, untargeted facial scraping, emotion recognition at work and in schools, and real-time remote biometric identification for law enforcement. None describes anonymous counting, which involves no biometric data. Camera systems that do infer demographics from faces should be assessed against the Act’s biometric provisions as the high-risk rules phase in (timeline).
United States
There is no federal privacy statute covering counting. State biometric laws — Illinois BIPA (the only one with a private right of action, and the source of most litigation), Texas CUBI, Washington RCW 19.375, and Colorado’s biometric amendments in force since 1 July 2025 — regulate biometric identifiers: face geometry, fingerprints, voiceprints (2025 litigation review). A counter that never captures a face has nothing to register. Face-based age and gender analytics, by contrast, is exactly what these laws were written for; if you want demographics in the U.S., get legal advice first.
Canada and Quebec
PIPEDA federally and Quebec’s Law 25 provincially govern personal information, with Law 25 carrying penalties of up to $10 million or 2% of worldwide revenue. Quebec’s regulator, the Commission d’accès à l’information, takes the strictest line in North America on biometrics — it refused a grocery chain’s proposed face-recognition database for shoplifting in 2025 (Torys) — and its 2024 anonymisation regulation treats aggregated statistics as the reliable form of anonymous data (Fasken). Hourly totals from a sensor that stored no image are precisely that. We are a Quebec company, and this is the environment our sensors were designed in.
Practical checklistPrefer sensors that never form an image, or that reduce images to depth on the device. Leave video recording off except for a dated audit. Do not connect Wi-Fi sniffing “for dwell time” without legal review. Report counts in intervals (15, 30 or 60 minutes) rather than as individual events. If you re-use CCTV, treat it as CCTV. And ask any vendor two questions in writing: what leaves the sensor, and what is stored where.
Matching sensor to door
Which people counter for which entrance?
The right technology is decided by the doorway, not the brand. Width, traffic density, door type, ceiling height, lighting and whether power can reach the entrance settle it. This is the same logic we apply on our own demo calls, with our products named where they fit and other technologies named where they fit better.
| Your entrance | Best-fit technology | Why | Our pick |
|---|---|---|---|
| Single or double door, outward-opening or none, low-to-medium traffic, people mostly single file | Horizontal infrared beam, wireless | Cheapest, no cable, minutes to install; two abreast counts as one, which at this traffic level is a small and consistent error | PEARL, $499.95, $0 software |
| Busy entrance, groups walking abreast, families with carts and strollers | 3D stereo (or ToF) | Separates people by head, filters by height, counts carts separately; needs one PoE cable | 3D Scope II LC, $1,149.95 |
| Wide mall lease line or open storefront (4–8 m) | 3D stereo, one or several stitched | One sensor covers up to 8 × 8 m at height; two stitched cover 8 × 22.8 m; beams are blocked by displays and people | 3D Scope II LC, multi-sensor |
| Frameless glass doors, automatic sliding doors with infrared safety sensors, inward-opening doors | Overhead sensor of any kind | Nowhere to mount a beam, the door swing breaks it, and door safety curtains interfere with infrared beams (our installation notes) | 3D Scope II LC |
| Very dark entrance (under ~3 lux), or an outdoor gate | Thermal, or an outdoor-rated 3D sensor | Thermal needs no light at all; stereo needs some. Outdoor stereo units are sealed and heated | 3D Scope II LC Outdoor, $1,650, or a thermal vendor |
| No network, no IT department, no cable possible | Battery beam over Wi-Fi, or battery thermal over cellular | The only two genuinely wire-free options in the market; the trade is battery maintenance and a monthly plan | PEARL on your 2.4 GHz Wi-Fi; a battery-and-cellular thermal counter if you have no Wi-Fi at all |
| Ceilings above 6 m — atriums, concourses, transit halls | High-mount 3D stereo | Our 3D Scope covers up to 6 m mounting height; specialist high-mount units are rated to 20–25 m | Not us — a high-ceiling specialist (named in the comparison) |
| You need age and gender, not just a count | AI video with face analysis | Requires real images and biometric processing; different legal footing in every jurisdiction above | Not us — an AI-video vendor that publishes demographic accuracy figures |
| Staff cross the entrance constantly (jewellers, boutiques, showrooms) | 3D stereo with staff exclusion | UWB tags worn by staff let the sensor ignore them without identifying anyone (how it works) | 3D Scope II LC + Staff Exclusion Kit |
Buying
Ten questions to ask before you buy a people counter
An update of the checklist we first published in 2020. The questions have not changed much; the answers vendors give have.
What is the guaranteed minimum accuracy, and how do you verify it?
“Up to” is marketing. A floor, a written guarantee period and a video-audit method are commitments. If the answer is “our sensors are 99% accurate”, ask what happens when yours is 91%.
What exactly does the sensor capture, transmit and store?
Get it in writing: image or no image, video on or off by default, what is kept on the device and for how long, and where the counts are hosted. This one answer settles most of your privacy obligations.
Does power have to reach the door, and who pays for that?
Battery sensors need none. Every PoE sensor needs a cable to the entrance — usually the furthest point from the comms room. Get the run quoted before you compare prices.
What is the total cost over five years, with everything?
Hardware, software, cable, installation, support contract, shipping and duties. Ask whether the software keeps working if you stop paying, and whether hardware is owned or rented.
Can it handle my door?
Width, height, door type, sunlight, traffic density, groups, carts. A vendor who asks you these questions before quoting is behaving well; one who quotes without asking is guessing.
How does it exclude staff, children and U-turns — and can I choose?
Height filtering, U-turn logic and staff exclusion are what make a count mean “customers”. Make sure the rules are configurable and that you know which rules are on.
Does it connect to my point of sale, and which systems are named?
Conversion rate needs transactions beside traffic. Vendors who name their integrations can be checked; “integrates with major POS systems” cannot. A software tier that computes conversion from manually entered daily transactions is a useful fallback.
What happens when it stops counting?
Sensors go offline, batteries die, someone hangs a sign over a beam. Ask how you are alerted, how fast, and whether the device buffers counts through an outage.
How do I buy it, and how long until it counts?
Card checkout and a courier, or a discovery call, a site survey and a contract? Both are legitimate; they suit very different buyers. Ask for the sequence, not the sales pitch.
Can I read the documentation and talk to support before I am a customer?
A public help centre and a demo call that answers hard questions are the best predictor of what support will be like when you own the thing.
Practicalities
Installing a people counter: heights, power, network and how many you need
Mounting height
Beam, adults only54 in (137 cm) — chest height. It filters children, carts and strollers, and it is above the height at which a swinging arm or a hand pushing a door would break the beam a second time. There is science behind what looks like a random number.
Beam, everyone24 in (61 cm) to include children — useful in libraries and museums where a child is a visitor.
Overhead 3D / ToFTypically 2.2–6 m (7–20 ft) ceiling height for standard overhead units; coverage width grows with height. Our 3D Scope II LC reaches up to 6 m and covers up to 8 × 8 m (coverage chart). Above 6 m you need a high-mount unit.
Overhead thermalTypically 2.5–4.5 m; narrower coverage per unit than stereo.
Placement rules that prevent most problems
- Mount overhead sensors 1–2 ft inside the door’s full swing, so the door itself never enters the count zone.
- Keep the count line clear of displays, greeters and queues; people who stop on the line are the main source of double counts.
- Avoid mounting over stairs, escalators, vestibules or revolving-door mechanisms; count on the flat floor just inside instead.
- Nothing may hang in the sensor’s view: exit signs, pendant lights, banners, plants, balloons.
- Beams need a clear line of sight and a place to fix both units; frameless glass shopfronts are the classic impossible case.
- Give a stereo camera at least 3 lux (we recommend 6); give a beam receiver shade from direct sun.
Power and network
BatteryWireless beams run 6–9 months on AA cells when set to count only during business hours; the trade-off is a maintenance routine and the risk of a dead pair nobody noticed. Battery thermal units with cellular run about two years.
PoEOne Cat5e/Cat6 cable carries power and data to a 3D or ToF sensor, typically under 8 W. An injector at the switch end means no outlet is needed at the door.
Wi-FiBattery beams use 2.4 GHz; check that your network offers a 2.4 GHz SSID with WPA2 and no captive login page — 5 GHz-only and enterprise-login networks are the usual snag.
CellularSome sensors carry their own SIM. Zero IT involvement, at the price of a monthly plan you cannot cancel without losing connectivity.
How many sensors?
One per public entrance, as a rule — a count with a door missing is a ratio with a hole in it. A staff or delivery door need not be counted unless customers use it. Entrances wider than a single sensor’s coverage take two or more stitched units working as one. Add a sensor facing the street or corridor if you want capture rate; add threshold sensors inside for department visits. Then, before going live, walk the line yourself twenty times in both directions, alone and in pairs, and check the totals. Our step-by-step implementation guide covers goals, placement, data handling and reporting cadence.
Vocabulary
People-counting glossary
- Traffic / footfall
- The number of people entering a defined space in a defined period. “Footfall” is the British and European term, “foot traffic” the North American one; they mean the same thing.
- Count line
- The virtual line a person must cross to be counted. Overhead sensors can have several, each with its own direction and rules.
- Directional counting
- Recording IN and OUT separately. Needed for occupancy; optional for traffic analysis, where the two are nearly symmetrical.
- Conversion rate (closing ratio, % sold)
- Transactions divided by traffic for the same period. The single metric a people counter exists to make possible.
- Retail Equation
- Traffic × conversion rate × average sale = total sales. Explains where growth came from, or did not.
- Capture rate
- Entries divided by passers-by outside. Measures the pull of a storefront or window; needs a second, outward-facing sensor.
- Occupancy
- People currently inside: cumulative IN minus cumulative OUT since the last reset. Accumulates sensor error; needs daily resets.
- Dwell time
- How long people stay. Estimated by tracking systems; not something a door counter measures.
- U-turn filtering
- Ignoring a person who crosses the line and immediately crosses back, so a greeter or a step inside to check the weather is not a visit.
- Height filtering
- Ignoring objects below a set height, to exclude children, carts and strollers. Done by mounting height for beams and by measured height for 3D sensors.
- Staff exclusion
- Removing employees from customer counts — today by an ultra-wideband tag they carry; historically by a separate staff door or a manual deduction.
- Group counting
- Counting people who enter close together as one “shopping unit”. Raises the apparent conversion rate; based on proximity, not on who is actually shopping together. Treat with caution.
- Stereo vision / 3D
- Two lenses whose slightly different views let the sensor compute the height of everything in the scene. “3D” and “stereo” are used interchangeably in this industry.
- Time-of-flight (ToF)
- A depth sensor that times infrared pulses to the floor. Produces geometry with no colour image at all.
- PoE
- Power over Ethernet: one network cable carries both power and data to an overhead sensor.
- Break-beam / horizontal beam
- An infrared transmitter and receiver facing each other across a doorway. A count is registered when the beam is interrupted.
- Accuracy vs consistency
- Accuracy is how close a count is to the truth; consistency is whether the error is the same every day. Consistency is what makes trends and ratios trustworthy.
Common questions
People counters: questions people actually ask
What is a people counter and how does it work?
A people counter is a sensor installed at an entrance that detects each person crossing a line and records the crossing with a timestamp and, usually, a direction. Depending on the technology it detects a broken infrared beam, a warm shape on a thermal image, a head-height object on a 3D depth map or a distance change measured by infrared pulses. The counts are sent to software that totals them by hour and day and compares them with sales and staffing.
How accurate are people counters?
Properly installed 3D, time-of-flight and thermal counters count 95–99% of people; infrared beam counters count about 95–98% of single-file adults and miss people walking abreast. Treat “up to 99%” as a ceiling; ask for the guaranteed floor and the verification method. Consistency — the same error every day — matters more than the last percentage point, because it is what makes trends and conversion rates trustworthy.
How much does a people counter cost?
In 2026, wireless beam sensors cost roughly $300–$500, 3D stereo and time-of-flight sensors $300–$1,200 (most between $650 and $1,200), and outdoor or enterprise units up to about $2,000, plus software from $0 to about $150 per location per month. If the sensor needs power at the door, add $150–$400 for a cable run. Our PEARL is $499.95 with a free software plan; our 3D Scope II LC is $1,149.95. Several large vendors publish no prices at all.
Do people counters record video or identify people?
Purpose-built counters do not. Beam, thermal and time-of-flight sensors cannot form a recognisable image at all, and good 3D stereo counters reduce their frames to a depth map on the device, transmitting only counts. Some camera-based counters can record short video for installation checks or accuracy audits; that should be off by default and used only on request. Counting systems that infer age or gender from faces are a different category and are regulated as biometric processing.
Are people counters legal under GDPR and privacy law?
Yes. A counter that stores no images and identifies nobody produces anonymous statistics, which are not personal data under GDPR, PIPEDA or Quebec’s Law 25, and it collects no biometric identifier under Illinois BIPA, Texas CUBI or similar state laws. The EU AI Act’s prohibited practices, in force since 2 February 2025, target biometric categorisation and identification, not anonymous counting. Camera systems that keep video or analyse faces do fall under CCTV and biometric rules and need proper assessment.
What is the difference between a people counter and a footfall counter?
None. “Footfall counter” is the usual term in the United Kingdom and Europe; “people counter”, “foot traffic counter”, “door counter” and “store traffic counter” are more common in North America. Libraries say “patron counter” and churches “attendance counter”. In French the device is a compteur de personnes or compteur d’achalandage. They all measure traffic: people entering a space over time.
Can I use my existing CCTV cameras to count people?
Sometimes, but rarely well. Security cameras are mounted at an angle to see faces, which is the wrong geometry for counting overlapping people; they struggle with shadows, groups and lighting changes, and using them for analytics brings the full CCTV rulebook on signage, retention and access. Dedicated overhead counters are mounted straight down, see height rather than faces, and are cheaper than the integration work usually is.
How do I calculate retail conversion rate?
Conversion rate = transactions ÷ store traffic for the same period. If 200 people entered on Tuesday and 30 bought, Tuesday’s conversion rate was 15%. Use IN counts (or total beam counts divided by two) as traffic, keep the definition constant, and compare like with like: hour to hour, Saturday to Saturday. Independent benchmarks put typical in-store conversion at 15–30% for specialty retail, 20–40% for grocery and 10–20% for big-box formats, and stores that start counting usually find themselves between 12% and 20% — but your own trend matters more than any benchmark.
Is manual counting with a clicker accurate enough?
No. In the one peer-reviewed comparison we know of, manual counts with tally sheets or clickers undercounted pedestrians by 8–25% against video of the same crossings, with error worst at the start and end of each session. A person also costs more per week in wages than most sensors cost outright. Manual counting is still useful for a short video audit of a sensor, or a one-off count where nothing can be installed.
Do I need directional (IN and OUT) counting?
Only if you need live occupancy. For traffic analysis, staffing and conversion, IN and OUT are nearly symmetrical over any period and directional counts add cost without changing decisions; count everyone and divide by two, or count IN only, and be consistent. Occupancy is different: it needs IN minus OUT from the most accurate sensor you can afford on every entrance, plus a daily reset, because every counting error accumulates through the day.
How do I exclude staff from the count?
Three ways, in order of accuracy: give staff a separate door that is not counted; use a 3D sensor with staff exclusion, where employees carry a small ultra-wideband tag that the sensor ignores without identifying anyone; or estimate a daily staff-crossing deduction and apply it consistently. For jewellers, boutiques and showrooms where staff cross the entrance constantly, tag-based exclusion is the only method that makes a conversion rate meaningful.
What is the best people counter for a small business?
For a single store with a standard door, a wireless infrared beam counter with a free software tier is usually the best value: no cable, no contractor, minutes to install, and a consistent count that is accurate enough for trends and conversion. Ours is the PEARL at $499.95. If your entrance is wide, busy with groups, glass-fronted or automatic, an overhead 3D sensor is worth the extra cost and the one cable. Nine systems are compared on published prices in our buyer’s guide.
Can a people counter work without Wi-Fi or a network cable?
Yes, two ways. Battery-powered beam counters connect over your existing 2.4 GHz Wi-Fi and need no cable at all. Battery thermal counters with a built-in cellular modem need neither Wi-Fi nor cable, at the cost of a monthly plan. Every 3D stereo and time-of-flight sensor on the market is powered over Ethernet and needs one cable run to the door.
Which people counter should a library or museum use?
A single-branch library with a standard entrance is well served by a wireless beam counter mounted at 24 inches so that children are counted as visitors, feeding software that keeps the full year of hourly data needed for the IMLS Public Libraries Survey. Multi-entrance buildings, wide lobbies, high ceilings and sites that must separate staff from patrons call for overhead 3D sensors. Either way, the point is a continuous count that replaces the “typical week times 52” estimate IMLS itself has flagged as unreliable.
Since 1972
The two counters we make
We sell one sensor of each kind that this guide recommends for ordinary doors: a wireless beam for standard entrances and a 3D stereo camera for everything harder. Both ship free to the U.S. and Canada, come with a two-year warranty and a 30-day money-back guarantee, and run on software from $0 a month.
PEARL — wireless people counter Ours
Infrared beam pair on AA batteries over 2.4 GHz Wi-Fi. Peel-and-stick, minutes to install, in/out detection, mount at 54 in for adults or 24 in for everyone. Upwards of 98% accuracy on adult traffic; covers doors up to 15 ft.
USD $499.95 · $449.95 with a 1-year plan · software from $0
3D Scope II LC — 3D people counting camera Ours
Stereoscopic 3D with on-device AI and HDR. Separates groups, filters children and carts, U-turn logic, staff exclusion, wheelchairs, multiple count lines, video audit on request. 99% rated, 95% floor guaranteed for two years. One PoE cable; covers up to 8 × 8 m.
USD $1,149.95 · $999.95 with a 1-year plan · outdoor model $1,650
~128 verified customer reviews across G2, Capterra and SourceForge (read them). Customers include City Furniture (a partner for more than ten years), The Body Shop Canada, La Vie en Rose, AM Retail Group and Furniture Mart USA (case studies).
Check our work
Sources
Claims about our own products come from our product pages, pricing page and help centre, all linked in place. Claims about the market and the law come from the independent sources below, checked on September 9, 2026. No competitor is quoted or linked on this page; named comparisons live on the comparison page.
Accuracy and manual counting
- Diogenes, Greene-Roesel, Arnold & Ragland (2007), Pedestrian Counting Methods at Intersections: A Comparative Study, Transportation Research Record 2002, pp. 26–30 — manual sheet and clicker counts undercounted by 8–25% vs video
- SMS Storetraffic — counter auditing policy (30-minute video audits)
- Wikipedia — People counter (history; 1988 Meadowhall installation; generational accuracy figures)
Wages and cost
- U.S. Bureau of Labor Statistics — Retail Sales Workers, median $17.03/hour, May 2025
- Government of Canada — federal minimum wage $18.15, April 1, 2026
- SMS Storetraffic — nine systems compared (published prices, five-year totals, cable-run rates)
Conversion benchmarks and market
- Shopify — Retail Conversion Rate (updated March 2026; in-store benchmarks by format)
- SMS Storetraffic — How to understand and grow your retail store sales (12–20% first-count observation; Larstan Consulting survey of 117 chain managers)
- SMS Storetraffic — Why you need a customer counting system (closing ratios; the 2% to 7% apparel case)
- U.S. Census Bureau — Quarterly Retail E-Commerce Sales (e-commerce 17.1% of total retail sales, Q2 2026)
- ICSC & McKinsey & Company — Shopping in the Age of AI: Redefining Stores for a New Era (April 2026)
- Mordor Intelligence — People Counting System Market; Fortune Business Insights
Privacy and law
- EU AI Act, Article 5 — prohibited practices; implementation timeline
- GDPR Recital 26 — anonymous information
- Privacy World — 2025 biometric privacy litigation review
- Torys — Quebec CAI on biometric information; Fasken — anonymisation regulation under Law 25
- SMS Storetraffic — 3D Scope II data-privacy recommendations
- Apple — private Wi-Fi addresses (why Wi-Fi counting declined)
Installation and specifications
How this guide was made
SMS Storetraffic publishes and maintains this page. It consolidates and updates six articles we published between 2016 and 2024 — the technology series, the manual-versus-automated piece and two introductions — and adds what changed: edge AI, time-of-flight, the collapse of Wi-Fi counting, the EU AI Act, and 2026 prices. The author has worked in automated people counting since 2006. We sell two of the products mentioned and say so wherever they appear; where another technology or vendor is the better fit, the guide says that too. Corrections are welcome at our contact page and are applied at the next quarterly review.