X5 Retail Group implemented the system of demand forecasting in Intersection

2020: Forecasting system implementation

On June 8, 2020 X5 Retail Group, the multiformat retail company of Russia, announced demand forecasting system implementation on the basis of machine learning in retail chain stores Intersection. The solution is completely developed in directorate according to Big Data X5 Retail Group, it allows to increase efficiency of advertizing campaigns and sales.

X5 develops own products of demand forecasting since 2019 on the basis of advanced technology of predictive analytics. The module on forecast calculation of amounts promo was tested in retail chain stores Intersection in March-April, 2020. As a result of the pilot growth of a gross income and efficiency of accomplishment of the plan of shop, reduction of trade inventories and write-offs is recorded. Own algorithms BigData X5 allow to reach the high accuracy of forecasting (more than 70%). The pilot was carried out by a technique And / V-testirovaniya.

For June 8, 2020 in a system 5 commodity categories, including in such positions, difficult for management, as fresh goods constantly are calculated with short terms of the validity. A system analyzes data from all 846 supermarkets of Perekryostok network. Until the end of 2020 the solution will be replicated on all categories.

The created forecasting model considers about 200 factors affecting demand. Among them price elasticity, traffic, assortment matrix, sales, advertizing activities, remaining balance, competitive environment, etc. In a system the analytics of sales by checks for determination of lack of goods on the shelf on hours and also recovery of a sales history is conducted that increases model forecast accuracy. Important distinctive feature of a system is daily updating of the forecast that significantly increases availability on the shelf and flexibility of all supply chain.

In reality of a post-pandemic digital transformation will become a key factor of success of business. A basis of this process are Big Data which directly affect process performance. The created system of demand forecasting in retail chain stores Intersection is developed by X5 command on the basis of machine learning technology, it calculates the calculation of goods, predicts load of shop, increases model accuracy on the basis of hundreds of factors that in, the final account, positively influences all process of a goods turnover, comments Anton Mironenkov, the director of Big Data X5 Retail Group.

In June the retail chain stores Intersection will start the pilot of the second system module designed especially for prediction of regular demand. Complete product introduction will come to the end in the first half of the year 2021.

The Accenture company acted as the partner in integration of a system into the operating IT architecture of network.