Indivd
•
•
Intro
Every one of five pet specialty stores in Northern Europe achieved at least 98.33% absolute people-counting accuracy, checked against a manual review of 2,562 entries and exits.
A pet specialty retailer in Northern Europe had Indivd's people counting checked at five stores in three countries. Every store recorded at least 98.33% absolute accuracy when automated counts were compared with manual video review. The QA reports were issued in September and October 2026.
The headline percentage is the lowest store-level result across the five locations. It describes counting accuracy during the reviewed periods.
The retailer chose the review periods at every store.
The stores
Reliable visitor counts behind store comparisons
A customer walks in with a dog on a lead and leaves carrying a sack of food. In pet retail, people, pets and bulky purchases can share the same doorway. The retailer needs a reliable count of the people visiting the store.
Pet retailers use visitor data to understand how busy each location is, when staff are needed and how many visits turn into purchases.
Those decisions depend on the count at the entrance. If one store undercounts visitors, its conversion rate can look higher without any extra sales. If another overcounts, its conversion can look weaker.
Across three countries, differences in counting accuracy can also affect comparisons between stores. Checking each location gives the retailer evidence about the quality of the numbers behind those comparisons.
The check
Five stores assessed using one standard
Reviewers examined masked video of real store activity and compared entries and exits frame by frame with Indivd's automated counts.
The verification covered:
Five pet specialty stores in three countries
2,562 entries and exits checked against video
67 review hours, added together across the counting lines
14 counting lines on six cameras
Entries and exits are counted separately. A person who enters and later leaves contributes two counts to the sample.
Each store's assessed counting lines were evaluated together. Reporting the lowest store-level result means stronger results elsewhere cannot raise the headline above the minimum.
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.
Consider an illustrative example: a manual review counts 1,000 entries and exits, while the system records 998. The totals differ by two, but the review finds six missed counts and four extra ones. Including all 10 errors gives 99.0% absolute accuracy.
What the number carries
A documented basis for comparing five stores
For the retailer, the value is a consistent way to assess the visitor counts used across five stores in three markets.
When teams compare conversion, they need to understand both transactions and visits. When they plan staffing, they need to understand when visitor demand builds. A documented counting check provides evidence about the footfall data used in both decisions.
Any improvement in conversion, staffing efficiency or sales would need to be measured separately.
What the check does give teams is a clearer starting point for investigating changes. If conversion rises or event-day traffic falls, they can examine the trading figures alongside documented evidence about counting performance.
Common questions
How accurate is people counting in pet stores?
In these five pet specialty stores, every store recorded at least 98.33% absolute people-counting accuracy when automated counts were compared with manual video review. This was the lowest store-level result across the tested locations, rather than an average.
How is people-counting accuracy verified?
Indivd's QA process compares automated entry and exit counts with frame-by-frame manual review of masked video. Reviewers record discrepancies and calculate absolute accuracy for the assessed sample. Reports identify the counting lines, sample size and review period.
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.
Can people-counting systems distinguish shoppers from pets?
A people-counting system used in pet retail needs to exclude animals from its visitor count. In these reviews, any animal counted as a person would have been recorded as an error, and every store still reached at least 98.33%. The reports do not give a separate figure for pet exclusion.
Why does accurate footfall matter for retail conversion?
Retail conversion is commonly calculated by dividing transactions by store visits. Undercounting visits makes reported conversion look higher; overcounting makes it look lower. Checking footfall accuracy helps retailers assess the visitor data behind their conversion figures and store comparisons.
What should pet retailers look for in a people-counting accuracy report?
Look for results from comparable store environments, a clearly defined accuracy measure, the number of entries and exits reviewed, sampling dates and the method used to check automated counts. Ask who selected the review periods.
Can a third party verify Indivd's counting accuracy?
Yes. Customers select the locations, periods and amount of testing, and can arrange verification without Indivd present.