Case study: Data-driven shelf management for Red Bull
Making sure that the shelves are filled with just the correct goods and that the warehouse is not overstocked is a logistical challenge, especially for big companies such as the producer of energy drinks Red Bull. Since 2018, when our cooperation with Red Bull started, we successfully implemented our Shelf Inspector solution in 15 stores of the retail chain Globus and 330 stores of the chain Albert.
The application helps brands coordinate the layout of specific goods on the shelves faster and more accurately so that the stock levels of various products are balanced. All you have to do is take a picture.
With the application, our merchandisers have more time left for more important tasks […] As the main benefit for the future, we see, in addition to saving time, also streamlining work, respectively price monitoring, where we can set various alerts in case the price of a particular product falls or rises above a certain limit.
Petra Diamond | the Lead of the marketing store
It's quite common for companies to have wildly inaccurate data about their out-of-shelf and out-of-stock products. Besides the need for monotonous counting of products on the shelves, which take too much time, there are also massive strategic assets for merchandising and secondary placement to improve the situation. The next problem lies in the loss of ROI (Return on investments), which, over time, can seriously affect the company's profits.
Currently, Shelf Inspector boasts 98.5% success in product recognition and is constantly improving. The application helps brands coordinate the layout of specific goods on the shelves faster and more accurately so that the stock levels of various products are balanced. All you have to do is take a picture of the stand or shelf in the store with a mobile phone. The application will analyse the data immediately, and track improvement in the refill scheme.
In practice, the sales representatives only need to take pictures of the shelves using the Shelf Inspector. Then, the application sends photos to a database stored in the Azure cloud, and the model runs neural network algorithms. As a result, within a few seconds, the system recognises what product is on the photo and reliably compares the results with the desired, most suitable situation. After that, Shelf Inspector designs the solution, which includes optimisation of the shelf layout.
- AI improves secondary compliance and merchandising priorities
- Product recognition saves up to 80% of the data collection time
- The app sends alerts about incorrect data input (e.g. blurry photos), which prevents misleading feedback and suggestions