Case Studies

The Effectiveness of 3D Cameras for Traffic Counting

Retail
The Effectiveness of 3D Cameras for Traffic Counting

Comparing the Performance of Dedicated 3D Cameras for Traffic Counting vs Existing (Legacy) Security Cameras

 

This customer, which offers lifestyle-oriented general merchandise and consumer products through a portfolio of global consumer brands, operates more than 500 stores in the United States, Canada, and Europe.

 

Business challenge:

Can existing security cameras be the source of reliable traffic counting data?

This customer was looking to obtain better traffic counting data and retail analytics capabilities. The full solution from Axper, featuring 3D cameras and store analysis, was chosen. However, in the initial phase, since the retailer’s existing security camera network deployed in its stores already provided some traffic counting functionality, they hypothesized that they could utilize this equipment while integrating only Axper’s traffic and store analysis.

Initially and for the first phase, all their data from the cameras were analyzed by Axper’s analytics software. With a baseline of comparable data established, 10 stores were selected to have Axper’s 3D cameras installed and ran the complete solution of cameras and analytics. From there it was easy to compare the before and after data and compare accuracy on a same store basis.

 

Observations:

Accuracy of Data and Impact on KPI’s

Oftentimes, retailers believe that their security cameras will be a reliable source of traffic data, but most often than not, these cameras are not positioned in the ideal location in the stores as their main purpose is security and theft prevention, not accurate traffic data.

2D cameras, although most often cheaper than 3D cameras are best suited for small and low traffic areas and typically lose accuracy as they often lose traffic count if visitors are too close to one another and may count a stroller as a person as they only generate a count without depth perception to differentiate and appropriately count what is being seen. Our accuracy test data is based on counting the number of visitors entering and exiting the store. Ideally, every visitor that enters should also exit, resulting in a 100% baseline accuracy.

In the case of our client, despite the cameras being positioned for the most part in appropriate locations near the store entrance, the difference in accuracy was significant. Across all stores, the 2D security cameras resulted in 87.4% accuracy with the lower per store accuracy was 71.5%. In contrast, Axper’s 3D cameras generated an overall 98.6% level of accuracy, with many stores at over 99% accuracy including the previously worst performing store that reached 99.6% accuracy!

The table below illustrates the impact of low data accuracy on when evaluating in-store performance KPI’s:

 

In this case, the sales performance on traffic (PoT) is inaccurate within a high variance of -12,6% in traffic data, leading to both inaccurate evaluations of staff productivity and customer potential.

The low accuracy in traffic data will also significantly impact staff allocation decisions and salary costs.

 

Basing staffing decisions on lower traffic data than actual will result in an under-staff store, lower levels of customer service and missed sales opportunity while an overestimated traffic will increase salary costs and decrease conversion per employee.

Data generated by security cameras also generated a high volume of support calls from store associates who believed their traffic was too high, resulting in additional costs from sending technicians to the store and the opportunity cost of losing data until technical issues are resolved.

 

Results

By comparing the before and after data on a same store basis, the gap in performance were clearly demonstrated and it was an easy decision for this retailer to move forward with the installation of Axper’s 3D cameras across its entire network of stores.

Once the data started flowing back to the store, the improvement in data quality resulted in strong buy-in from store associates. Uptime also improved drastically, also increasing the confidence in the value of the data and reducing the maintenance and repair cost that the older security cameras required.

Since Axper’s solution has been fully deployed, our customer has come to rely on and trust its store traffic data to make informed store management decisions in implementing programs and strategies.