M.Video-Eldorado introduces in-store video analytics

18.03.2021 14:48

M.Video-Eldorado Group (M.Video-Eldorado Group, the Company, or the Group; MOEX: MVID), Russia’s leading e-commerce and consumer electronics retailer and part of SAFMAR Group owned by Mikhail Gutseriev, expands its pool of digital retail technologies to improve customer service and effectively manage business processes. The Company is testing in-store human resources and customer service management based on video analytics data that is processed by a neural network in real time. The pilot project runs three working analysis scenarios: lone customer, queued checkouts, and store heat map.

The video analysis system was developed by M.Video-Eldorado’s data office. The data from in-store IP cameras is processed by a neural network based on YOLO, a solution to scan images for multiple object detection. The video stream is processed using cloud solutions and infrastructure based on Raspberry. This smart solution analyses the real-time data flow from the store based on pre-set patterns. It can distinguish employees from visitors and then superimposes their location data on the store layout.

The first working scenario for testing in-store video analytics was to help customers who tend to stand or move around alone for some time. The solution helps quickly identify such customers and sends a signal to the store chatbot, following which a free consultant approaches the customer and provides personalised assistance. This innovation contributes to better staff attention and customer service, resulting in a fivefold growth in employee engagement.

The neural network also analyses the number of visitors in the pick-up and checkout areas. If the limits are exceeded, consultants receive a message and take action to resolve the situation.

Heat map is another solution based on the neural network which is designed to analyse retail space and sales management. This solution builds a density distribution of store visitors by area, which helps understand customer behavioural patterns, assess the convenience of product placement, and choose locations for promo placement.

M.Video-Eldorado plans to run pilots in a number of its stores. Going forward, the solution will be deployed in more than 1,000 stores should its cost-effectiveness be proved.

Kirill Ivanov, Head of Data Office at M.Video-Eldorado.

“The retail infrastructure lies at the core of our business helping customers quickly access products, including online orders, test them on their own and get quality advice. As customer behavioural patterns evolve, we seek to maximise the digitalisation of retail processes, improve the customer experience, and create a personalised seamless experience for our buyers based on the One Retail model. For instance, we are successfully developing the mobile platform with integration of mobile apps on both consultant and customer sides and in-store contact-free services, while also leveraging predictive analytics to manage stocks and deliveries.

“We developed the video analytics solution from scratch in less than six months using our own team of graduates and students of the Higher School of Economics and without investing in expensive third-party out-of-the-box platforms. The data we obtain allows us to better understand our customers, their needs and habits and make the in-store analytics as rich as the one sourced from the website or the mobile app going forward. The behavioural analytics helps us bring customer interaction to a whole new level and meet their personal requirements, while also enhancing the efficiency and quality of our own business processes. In the near future, we plan to expand the capabilities of this tool by adding new working scenarios, such as the registration of group visits.”

As at the end of 2020, M.Video-Eldorado Group had 1,074 stores, with some 500 more stores of various formats expected to be opened over the next three years. 90% of the Group's turnover is related to in-store customer experience, including in-store shopping, pick-up of online orders and express delivery by taxi from the nearest store. Currently, 60% of our customers across Russia can receive their online orders within 24 hours, while more than a third of orders can be ready for pick-up within 15 minutes