Editors' pick 12 May 2022
The most common misconception about people counting accuracy
People who have decided to invest in people counters, are often driven by an urge to understand their customers better, to know who their store visitors are, to be able to analyse the store's conversion rate or to understand the customer journey for the people who visit their shop/mall.
To get answers to those questions and much more, you rely on the people counters being accurate and that you can trust the insights are correct.
The truth, however, is that the accuracy of a people counter can be poor, meaning it provides a false sense of security. It does not explain how well you can trust the insights.
The difference between accuracy and reliability
If you google people counters and go to the website of a random supplier, you will soon see that they're addressing the issue of accuracy.
When they refer to their accuracy, they talk about the test results from a "lab environment" and the error margin in detecting one person.
This lab environment is, however, not a good representation of how well it works in your shop or mall. The real environment contains lots of interference. Interference is everything that's unforeseen, but can occur. For example, some visitors walk outside the detection range, the light is different in different areas of the premises, two people may walk close to each other, a person has an odd outfit or may be carrying something, and so on.
There is a lot of interference that is not covered by the accuracy measure. Unlike reliability, which explains how well you can trust that the insights represent reality.
How is reliability measured?
Reliability is measured through empirical studies, beginning with installing the equipment in a randomly selected space. Then, applying a controlled method, you measure how well the equipment performs against a number of key performance indicators - KPIs.
Some of these KPIs are Visitor Detection Accuracy (consisting of about 20 KPIs), Occlusion Accuracy, Tracing Accuracy, Exclusion Accuracy, Detection Coverage, Transition Accuracy and more.
The data is then analysed, and the statistical significance is established. Once that's done, one can draw conclusions as to how reliable the insights from a people counter are.
But then, are reliable insights from people counters really that important?
The answer is probably - It depends. It depends on what you use your data for.
Do you use your insights to become more digitally mature, to make strategic decisions, to increase sales, to streamline marketing, to grow the profitability in your business, to create a more sustainable value chain?
If the answer is yes to one of those questions, then you must to be able to rely on the data. Or else, you'll always find yourself questioning the data collection process.
This is how reliable the insights from Indivd's people counters are
Our journey started in 2017 as a humanitarian, self-funded research project committed to creating better conditions for brick-and-mortar businesses.
This was partly because we had purchased some people counters and realised they did not really provide a solution to the problems at hand, but particularly because they did not address the issues around integrity.
Not only did brick-and-mortar businesses lack reliable insights, but they also didn't really have access to the kind of valuable insights that were - and are - crucial for them to secure their business’ future prosperity.
A 2019 study from Retail Academics Research Institute concludes that stores and sites lack 77.22% of the total insights they need in order to reach the digital maturity required in the industry today.
That's why we, for several years now, have researched and patented a new type of people counter that makes stores and sites to become sufficiently digitally mature to solve issues crucial for their future prosperity.
We continuously test and enhance the reliability of insights from our people counter. Apart from offering entirely unique insights, we also have the market's highest reliability.