M.Video - Behavioral cluster planning
Galleria Helps Russia’s M.Video Improve its Market Position and Increase Sales
Since its founding in 1993, M.Video has grown to become a leading European consumer electronics retailer – now operating 277 brand name stores in 121 Russian cities (as of June 30, 2012). The total space of M.Video stores accounts to 685,500 sq. m, the selling space - 508,500 sq. m.
The management of M.Video has set several goals to support this growth: to provide all of its customers with great value and the best in technology products and offer a superior shopping experience and provide the very best customer service.
By 2010, M.Video was operating a variety of assortment ranges in stores. To establish space considerations and assortments for its stores, M.Video used the average revenue by its total number of stores to analyse which products were selling and which were not – then took the information to determine the number of lines offered at each of its stores. M.Video’s category managers then decided the size of ranges and adopted a policy of standard pricing in most stores – which typically resulted in a package of standard promotions.
This process would often result in placing too many product lines and failing to provide more of what customers could actually buy. This created too much or too little space with inventory devoted to the wrong products.
Management came to realise that if they could group similar stores together and manage the assortment and merchandise differently, they could improve the stock mix, reduce inventories and provide a better assortment of products for customers at each of its stores – creating a major impact on sales and profitability.
Working with consultants from accounting firm Ernst & Young, M.Video searched for a company that had the experience and software capabilities to help management optimize the space in those stores that shared similar behaviors. Coupled with other work it was already doing on range assortment, M. Video set out to build the foundation for an end-to-end solution that could help it manage its complete supply chain process.
Enter Galleria and Behavioural Clustering Software
In January 2011, M.Video began working with Galleria, the leading provider of retailer and supplier category optimization solutions and consulting services known for its previous work with many of the world’s top retailers, including: A&P, Giant Eagle, One Stop, Safeway and Tesco, Galleria provides customer-focused solutions for clustering, automated assortment and space optimization supported by detailed analytics and reporting tools designed to meet the needs of retailers and CPG vendors and suppliers.
Galleria also developed a strong track record in helping its retail clients increase sales, enhance margins, reduce waste/mark-down and accelerate inventory turns.
The team at M.Video knew that a generic approach to store clustering would typically result in out of stocks, lost sales or drive excess or unwanted inventory in stores. They also realised that the resources required for that type of approach were time consuming, costly and wasteful.
However, with Galleria’s Behavioural Clustering Software, store clusters would be based on consumer buying behaviours at similar stores – ensuring that like-stores would each reflect the same space considerations for products and the same assortments. Galleria’s Behavioural Clustering solution involves a ‘three click approach,’ which includes selecting the required time series data, creating the Customer Decision Tree and then developing the store clusters. That information is then supported by demographic and geographic data within the software to help determine what type of stores are within the clusters. Reviewing the results in a graphical report format is an integral part of the solution however data can be transferred into the corporate data mart to support broader decision making.
Galleria’s Behavioural Clustering Solution enabled M.Video to identify variation and diversity within product and store attributes and analyse that information against shopper demographic and geographical data – in a process that takes hours in preference to weeks and does not require an army of PHDs’ to drive it.
Optimizing television sales
In February 2011, M.Video, Ernst Young and Galleria began to explore television sales throughout the chain of stores. The project included training for management and key employees, the installation of Galleria software, importing data and analysing the results.
From their work, seven behavioral clusters emerged – three “Mainstream,” two “in City” and two “Suburbs” clusters – each based on meaningful differences in main selling patterns. The solution took several television attributes into account, including screen size, technology, brand and price. Galleria used M.Video’s sales value metric for six previous months of sales data so the latest trends were reflected in the process.
After building a number of different customer decision trees and analysing in various priorities, Galleria concluded that the most accurate clusters to develop for the data included screen size, technology and brand. Although retail prices were left out of the calculations, since their attributes were very similar to screen size, prices were considered as part of the cluster analysis. In addition to the product data, Galleria also used store location and geographic information to identify the clusters.
Galleria then added a number of new attributes to the solution to enhance the final store cluster analysis: M.Video store information including store sales statistics and promotion participation.
Following further analysis and refinement, and taking into consideration the shares of various attributes, low sales and unusual trends within some stores, it was determined that M.Video’s locations could be aggregated into three store clusters – Advanced, Economy and Mainstream.
The new clusters now showed clear distinctions in purchasing patterns by screen size, technology and brand. It was also clear that each of the clusters had very different promotional participation levels.
With the new information provided by Behavioural Cluster Planner, it became very easy to summarise the new Advanced and Economy clusters.
Customers in the Advanced Cluster preferred newer, cutting-edge technologies, and were more likely to buy a small television as a second TV or as a present and were unwilling to buy secondary brands. Meanwhile, customers in the Economy Cluster preferred base models with an average size for the best deal and would wait for a price reduction before buying an expensive TV. They also preferred buying older models with clearance markdowns.
Following the final analysis, M.Video management discovered new opportunities for other quick wins in their business:
|Promotions||High Technology||Core Range|
|Customer Service||Delivery & Installation||Base Level|
|Stock Levels||High Technology||Core Range|
|In Store Features||Connectivity Zones||Clearance Zones|
Shaping assortment to fit customer behaviour leads to spike in sales
Through its work with Galleria, M.Video now has the ability to create store groups based on actual consumer buying behaviours – and it’s already experiencing positive results. Since the project was implemented, M.Video management reports that sales of televisions have increased in excess of 11 percent and stock turnover improved of 12 percent over the last year compared to the previous year.
“Galleria’s Behavioural Cluster Planning solution has enabled us to highlight variations across our store estate” said Konstantin Nechaev, Program Director at M.Video. “Armed with this insight, we’ve been able to tailor the stores more closely to the customer requirements. This ensures that we carry more lines that are appropriate to the needs of customers, while still providing a range of products that fall outside of these identified clusters.”
“With the Galleria solution we can now analyse demand patterns, draw conclusions and convert them into particular actions in a very short period of time,” Nechaev said. Galleria’s Behavioural Clustering Solution has put M. Video on a solid foundation to preserve and expand its dominant market share in the face of growing competition from other consumer electronics stores – and offer “the right item on the right shelves in the right quantities” for its customers – regardless of the store location.
As for the future, M.Video is already extending the Behavioural Clustering program into other product categories – including washing machines, refrigerators, cameras and laptops.
Through Galleria Behavioural Cluster Planning, M.Video now has the opportunity to take a whole new look at how it promotes, prices, merchandises and displays products by store cluster to increase sales and improve the customer experience.
"..we’ve been able to tailor the stores more closely to the customer requirements." Konstantin Nechaev, Program Director, M.Video