People counter accuracy describes how closely the number reported by a sensor matches the true number of people who walked through the entrance. It is usually expressed as a percentage, found by comparing the sensor count with a manual count over the same period.
Accuracy depends on the sensor type, the mounting height and position, the width of the doorway and the conditions. Groups walking shoulder to shoulder, prams, trolleys, bright sunlight on the floor, glass doors and people lingering in the doorway can all cause over-counts or under-counts. Overhead 3D sensors are generally designed to handle these better than simple infrared beams, but installation still matters.
Operators check accuracy when a sensor is first installed and again after any change to the entrance, lighting or fit-out. A typical check is to have someone tally entries by hand for an hour or two at a busy time and compare the figure with the sensor. Repeating the check seasonally, or after the door hardware or signage changes, catches problems before they skew months of reports.
Be cautious with supplier claims. Accuracy percentages are usually measured under good conditions and may not match your door. Ask how a figure was tested, and verify it on site. A small, consistent error is less harmful than an unpredictable one, because trends and comparisons still hold.
Every downstream metric, from conversion rate to staffing, relies on the visitor count, so a poor count quietly distorts decisions.