From the years 2020 to 2022, online retail received a significant boost to its already good growth due to the global pandemic, as shown in the annual report of the Federal Statistical Office Germany as well as in the HDE Online Monitor 2021. Among other things, providers of online shops benefit greatly from the ability to evaluate large amounts of profound and individual customer data. In order to provide a similar advantage to brick-and-mortar retailers, Intenta GmbH has expanded its sensor portfolio with the S1000 people counter, a specialist for people counting with the goal of supporting traditional retailers in obtaining com-prehensive customer data.
Key performance indicators of customer footfall
In order to measure the performance of stores or shopping centers, 3D smart sensors are used to gather counting results and tracking data. These are the basis of the following key performance indicators and activities:
- Customer occupancy of a store in real time
- Success of marketing activities and sales promotions
- Comparison of the customer volumes in different stores
- Staff planning based on customer footfall
- Conversion rate
- Hot and cold area mapping
With knowledge of these statistical parameters, you can obtain an accurate picture of customer behavior in the store.
Stereovision technology for accurate counting results
In order to generate reliable information, it is essential that the counting data is as accurate as possible. In order to achieve this precision all Intenta 3D-Smart sensors use advanced stereo-vision technology. This works by observing an object or a person through two cameras positioned at a fixed base distance. This allows one to obtain a depth information similar to human vision from which a digital 3D model can be generated.
By doing so, significantly more accurate counting results can be produced than, for example, with 2D-video technology. Since all image data is processed in the sensor, only anonymous data is transferred for further analysis. For the proof of counting accuracy, validation data can be recorded with the sensors and later evaluated in a separate tool.
