Case Study: Data-driven shelf management for Nestlé
While developing and implementing our solution that helps merchandisers eliminate out-of-stock products, we have worked with a leading retail chain in Slovakia with hundreds of outlets. One of the benefits of our Shelf Inspector solution is that it can recognize SKUs (stock-keeping units) in packs and boxes and quickly identifies products bottled in glass. As a result, the application optimizes merchandisers’ and sales representatives’ work, and saves time and money.
We faced multiple non-digitalized planograms with more than 500 SKUs arranged into five categories during our work on this project. All we had were photos of the shelves where many products looked similar. Another challenge was associated with the redesign of 15% of the portfolio, which is a relatively significant number concerning hundreds of stores.
To solve the existing problems, we combined neural networks to detect the product design and size. It allows the merchandiser or sales representative to log into the Shelf Inspector app, and then they only need to choose the category and the products. The application leads the user through taking a photo and displaying the required goods. It will show crops of the picture with a semi-transparent mask over the peak of the products – delimitating the different kinds of packaging and sort them by size.
There is an in-app tool dedicated to monitoring the out-of-stock and out-of-shelf products, and currently, there is an automated PowerBI reporting that runs three times a week.
We created the app with the help of SaaS technology, which means it is a cloud service, and all information once saved in the system is stored in a cloud and easily accessible from any device.
- Product recognition makes it easier to monitor out-of-shelf, out-of-stock, and pricing
- Logistics employees or store manager will be notified about low stocks
- AI improves planogram compliance
- Recreating planograms in TT and logistics planning leads to fewer out-of-stock situations