Nov 3, 2026 · 4 min read · By XenReality team
How Footfall Counting Works: A Practical Guide for Retailers
Footfall is the starting point for judging store performance. Here is how camera-based counting works, what it measures and what to plan before you start.

What footfall counting measures
Footfall is the number of people who enter a store, a floor or an area in a given period. It is the denominator for almost every retail question: conversion rate, sales per visitor, staffing levels and the return on a promotion all depend on knowing how many people actually came.
How camera-based counting works
A camera is mounted above or facing an entrance. Software detects each person in the video and follows them as they move. When someone crosses a line you have defined, or enters a zone, they are counted. Because the system analyses video rather than infrared beams, it can tell people from trolleys, avoid counting the same person twice and separate entries from exits.
- Detect people in each frame of the video
- Follow each person from frame to frame
- Count crossings of a virtual line, or entries to a zone
- Summarise the counts by hour, day and location
What else the same video can tell you
- Dwell time: how long visitors stay in a zone, which shows where interest is
- Heatmaps: which parts of the space are busy and which are overlooked
- Queues and occupancy: how many people are waiting or present at once
- Peak hours: when the space is busiest, to match staffing to demand
What to plan before you start
- Camera position: high, angled cameras reduce people blocking each other at busy entrances
- Lighting: stable lighting at the entrance gives the most reliable counts
- Definitions: decide whether staff, children or people passing by are counted
- Questions first: know what decisions the numbers are meant to support
- Privacy: counting does not need to identify anyone, so choose a setup that does not store faces
Start with a pilot
Accuracy depends on your entrance, your lighting and how crowded it gets. A short pilot on your own cameras shows the accuracy you can expect before you commit, and it often reveals a better camera position at almost no cost.
Back to blog