Indivd
•
•
Intro
A two-level fashion store recorded 98.29% absolute people-counting accuracy across 52 lines, including till areas, fitting rooms and departments.
A fashion retailer in Northern Europe had Indivd’s people counting checked across one of its stores, using samples collected on 31 August and 1 September 2026.
The assessment extended beyond entrance counting. It covered 52 counting lines on 19 cameras, including lines associated with till areas, fitting rooms, departments, lifts and stairs. Manual video review recorded 148 counting errors across 8,675 counted movements, giving a combined absolute accuracy of 98.29%.
For the retailer, the check provides evidence about the counts used to compare activity within the store and inform staffing and layout decisions.
The store
One store, measured beyond the door
A large fashion store asks more of its visitor data than how many people came in. Teams want to know how many people moved to the lower level, how much traffic reached the fitting rooms and the till areas, and which departments drew the most activity.
Each of those questions needs its own count. An entrance figure shows the size of the store’s traffic. It cannot show how that traffic spreads across levels, departments and service points.
Movement inside a store is also harder to measure than an entrance. Visitors use lifts in groups, double back on stairs, browse at the edge of a department and cross the same line several times in one visit.
At this store, the assessment covered lines at the entrance, between the two levels, at fitting rooms on both levels, at till areas, and at the beauty, accessories and children’s departments.
The check
52 counting lines, assessed together
Indivd’s QA team examined masked video of real store activity and compared Indivd’s automated counts with manual counts, frame by frame.
The verification covered:
8,675 counted movements checked against video
52 counting lines on 19 cameras
176 review hours, added together across the counting lines
Two sampling dates, 31 August and 1 September 2026
What the 52 lines covered
Where | Lines |
|---|---|
Till areas | 18 |
Beauty department | 10 |
Children’s department | 8 |
Lifts | 4 |
Fitting rooms | 4 |
Entry lines | 2 |
Stairs | 2 |
Accessories | 2 |
Other internal line | 2 |
Total | 52 |
Each crossing of a counting line is one counted movement, in either direction. Because most lines sit inside the store, one shopper can contribute several counts during a single visit: entering, taking the lift, visiting a fitting room and passing a till area.
The complete set of lines was assessed together. The review recorded 148 counting errors across 8,675 counted movements, giving 98.29% absolute accuracy.
DEFINITION
Absolute accuracy is 100% minus the absolute error rate. Missed counts and extra counts are both errors, and errors in opposite directions cannot cancel each other out.
This assessment was carried out by Indivd’s QA team. As a general process, customers can choose where, when and how much to test, and can appoint a third party to carry out the review. Read more about how Indivd verifies counting accuracy.
In a separate customer-commissioned review, KPMG applied Indivd’s quality-assurance methodology and recorded approximately 99% absolute counting accuracy. That review was separate from this assessment. See the guide to testing people-counting accuracy claims.
What the number carries
A checked count beyond the entrance
For the retailer, the value is a documented check of the counts used to compare activity across departments, fitting room areas and the connections between floors.
Those counts feed practical questions. How many staff should be on the tills and at the fitting rooms at each hour? Does activity on the lower level justify its share of stock and space? Which departments draw more traffic after a layout change?
An error at an internal line travels into those answers. An undercount at a fitting room area can make it look quieter than it is. An overcount at a department can make it look busier than its sales would suggest.
The QA review addresses counting accuracy. Alongside sales and staffing data, these counts can help teams investigate where customer activity and staff coverage may be out of step.
The retailer can now compare activity across levels, departments and service points, with each count checked at the line where it was produced.
Common questions
What is zone-level people counting?
Zone-level counting places lines around areas inside a store, such as fitting rooms, till areas or departments, and counts people crossing in each direction. It shows how much traffic each area receives. It does not follow individual visitors between areas. In this store, 50 of the 52 tested lines sat inside the store, and all were checked against the same manual video review.
Can people counting measure movement between floors?
Yes. In this store, lines at the lifts and stairs between the two levels were part of the complete counting set. They were checked against the same manual video review as the entrance, fitting room and till area lines, and the combined result for all 52 lines was 98.29%.
Why count fitting rooms and till areas as well as the entrance?
The entrance shows how much traffic the store receives. Fitting room and till area lines show how much of that traffic reaches those service points. Set against transactions and staffing, those counts help teams see where activity builds up during the day. They do not show individual try-ons or purchases.
What does absolute accuracy mean in people counting?
Absolute accuracy is 100% minus the absolute error rate. Missed counts and extra counts are included separately, even when they offset each other in the final total. This makes errors visible that a comparison of totals alone could hide.
Why does one visit produce several counts?
Each counting line records crossings in its own location. A shopper who enters, takes the lift and visits a fitting room crosses several lines, so one visit can add several counted movements to the sample. The entrance lines alone show the number of store visits.
What should a retailer look for in a people-counting accuracy report?
Look for the tested locations and counting lines, sampling dates, number of movements reviewed, comparison method and definition of accuracy. Check whether the headline describes a combined result, an average or a minimum, and whether internal lines were tested as well as the entrance.