RussiaRetail and e-commerce2021 – ongoingMagnit
Promo forecasting and on-shelf availability monitoring
Magnit said big data and AI projects added 1.5 billion rubles of EBITDA in 2021
Problem
Magnit needed to estimate how many discounted items each store would sell and to spot empty or misplaced shelves quickly across thousands of small stores.
Approach
Magnit built its own promo forecasting system and used machine learning to evaluate past promotions. A data center analyzes sales in near real time and alerts store administrators by phone when shelves need restocking. In 2021 the availability system ran in 9,500 convenience stores.
Results
Takeaway
The EBITDA figure covers both projects together, so the gain from each is not known. Magnit said it planned to expand them in 2022.
Categories
- Size
- Large enterprise
- Industry
- Retail and e-commerce
- Business function
- Production and qualityMarketing and sales
- Innovation type
- Process innovation
- AI technology
- Predictive analytics and ML
- Adoption
- Built in-house
- Evidence
- Company's own claim
Sources
Source summaries
Retail.ru reports Magnit's statements on its 2021 big data and AI work. It gives the EBITDA gain, the forecast accuracy rise and the number of stores with shelf availability monitoring. Neural-network promo prediction and personal offers were described as still in development or planned.
Full details are in the original sources.
