Over the last years, X5 Retail Group, a leading Russian food retailer listed on the London and Moscow stock exchanges, has undertaken a process of deep digital transformation from procurement automation, to facial recognition, to video analytics. As Head of X5 Technologies, Anton Mironenkov, leads a team of some 2,500 professionals. In this interview with East-West Digital News, he shared insights on how X5 is conducting these changes both internally and via open innovation programs. He also commented on the way technology has been instrumental to make the group’s operations more resilient to the current pandemic.
How would you characterize today’s key Big Data and AI trends in retail?
If taking the example of X5, this major retail group started looking more and more like a tech company in recent years. Big Data and machine learning, video analytics of store shelves, facial recognition, automated procurement, and digital voice assistants have all become part of our operations. One of the most important focus areas of our digital transformation strategy is the creation of ready-made analytics services based on the accumulated big data. We use big data both for internal use and for external monetization purposes and think of big data as one of the most important assets of our business.
X5 has accumulated big data analytics and coupled it with a service available to each of our 2,000 analysts. They use data on customer transactions, X5’s suppliers and their counterparties to make faster decisions. So, algorithms have become an integral part of X5’s data-driven culture, where all of our decisions are made based on data and advanced analytics. This is how we mitigate human error and speed up decision making.
The decisions based on this approach have become more informed and more accurate, as advanced methods of data analysis facilitate a sharper focus on the customer and greater operational efficiency. X5 collects terabytes of data daily and uses its own resources to develop analytical products and services to differentiate our offering from competitors. Needless to say, we pay great attention to data protection and have adopted global best practices in this area.
Proper use of big data to support business strategy requires constant experimenting, prototyping and extraction of new knowledge from our data sets. A number of our big data tools have already passed this stage and are now being routinely used in our activities.
What kind of technology tools do you use for assortment management?
We use there automated category analysis tools to ensure that our customers’ needs are fully met. This automated assortment management system helps us to form an offering in 28 product categories. New assortment matrices are developed based on our understanding of demand, the relevance of each product category and its share in total net retail revenue. This complex process is key to ensuring our customers can always find what they want at are stores, which we stock in the most efficient way possible.
What about marketing?
In this field, promotions better hit the mark if we use advanced analytics based on big data and takeaways of past promotions, which help us to make more accurate and rational decisions. This system will be developed further to predict regular demand based on predictive models, facilitating greater logistics efficiency through more accurate scheduling of delivery times and volumes. Customer Value Management aims to further create additional value for X5’s loyal customers, as well as maximise the value of each loyal customer to X5.
Looking back, one of the first digital products launched by X5 was the Automated Pricing System at Pyaterochka. It took us a year to developed and this summer, we rolled it up across all the regions of Russia where Pyaterochka operates. This system is set to increase our margins while maintaining retail sales volumes and customer traffic. While it is already in operation, this system is continuously enhanced — mathematical models and application interface are being improved and additional reporting is being introduced. This is a tool that helps to find a balance between profitability and price perception.
Did the Covid pandemic alter your digital transformation plans?
In this pandemic, all the digital transformation measures which were previously launched – at the end of 2018 – are proving essential. Big data has been at the core of many of our practices, including analysing and predicting consumer behaviour. Analytics systems monitoring our loyalty programmes have shown a 24% year-on-year increase in purchases with loyalty cards for X5 retail formats between September 2019 and August 2020.
The increase in customer spending has been a key growth factor. Pyaterochka made a major contribution to X5 Retail Group’s sales volumes, with Perekrestok also growing at an impressive rate. Online services have seen higher demand than in-store services even as quarantine measures are relaxed, while our overall share of customers is higher than before the lockdown began. The change in demand during the lockdown is primarily due to customers buying more items per shop at a higher average price.
Demand was also changing across categories during the lockdown; sales of alcohol – long-life groceries, home care and basic hygiene products were growing and selfcare products were seeing lower demand. As quarantine measures were relaxed, the opposite trends were in play and impulse purchases increased.
Even though the coronavirus hasn’t said its last word yet in Russia, customer behaviour is gradually normalising. As Covid measures are relaxed, the number of customers is going back to pre-lockdown levels, with people buying fewer goods per shop as they visit our stores more often.
Is your approach to Big Data and retail tech different from world practice? from your competitors in Russia?
Below are X5’s big data-based products, which are unique for the Russian market:
- Proprietary A/B testing platform. In Russia, five companies at most (digital ones, not retailers) have such platforms available. Worldwide, there are about 20 such companies (Google and Facebook among them). The A/B testing platform helps us verify the results of any pilot project or business process with maximum accuracy by simulating the environment and selecting identical factors as if creating an alternate reality to confirm the results of an experiment. In general, predictive methods based on machine learning can be used to assess whether promotions gained traction among customers, whether customer outflow slowed down, while also gauging price elasticity, calculating the effect on retail sales volumes or customer traffic.
- Own algorithms to determine pricing, assortment, procurement and increase retail sales volumes
- A unique platform for suppliers featuring accumulated data on sales and customer behaviour. Now every supplier can receive almost real-time analytics reports and cooperate with X5 even more efficiently. Dialog X5 is a multi-purpose platform for tackling business tasks, a set of digital tools with built-in analytics and logistics reporting, promotion targeting and a tender platform. X5 plans to integrate all services for suppliers into a single system based on the Dialog X5 platform.
- Recently we announced the launch of X5 ID, a universal account with access to all of X5’s retail services with a single login. No other Russian retailer has such a solution, based on best global practices. All of our store formats and business units will use X5 ID to create an integrated user experience, boost the usage frequency of our services, and save customers’ time.
- Big data as a pillar for store digitalisation, including face recognition and product freshness monitoring algorithms, video analytics of shelves and queues, etc.
What are the contributions of internal teams, and of open innovation, in X5’s work on Big Data?
Today, the majority of X5’s Big Data decisions are implemented by the X5 Technologies team, which brings together specialists with unique skill sets, even by global market standards.
The quality and availability of data are our key success factors, and X5 pays special attention to organising the process and culture of data management. We actively implement a data-driven management culture, and today this process involves all business divisions and teams. This data-driven management culture is also based on an approach of ‘democratising’ data for those it will benefit in our business.
X5 is part of the Retail Innovation Tech Alliance (RITA), an international startup scouting initiative launched by major Russian retailers. Six months after the first call, has this program produced results for X5?
The RITA project, initiated by X5, is a fundamentally different approach to innovation. Such an alliance is unprecedented not only for Russian retail but for the global market as a whole. We pool our efforts to carry out effective scouting around the world. This platform is a place where we will freely share experience, assess solutions and implementation results, so we can test new technology faster.
Often retailers face similar challenges and tackles the same tasks on their own. We have common goals – improving the customer experience, increasing operational efficiency, streamlining logistics and back office operations. We can find solutions on our own, but will make mistakes along the way, as some solutions turn out to be inoperable, hard to scale or adapt to existing business processes. Pooling our experiences will help all of us increase return on investment and implement viable solutions.
One more mission that RITA undertakes is to give a clear signal to the market about what the retail industry is looking for from startups. By joining our efforts, we aim to attract the best startups in our segment. Big businesses have no problems attracting local startups, but the situation is different when you consider startups from other parts of the world. A startup with a high-quality business solution, whether it is developed in China, Israel or a country in Europe, can come to any market and will likely choose their location on the highest chance for monetizing its product. In that respect, RITA represents a much greater buying power, becoming more attractive to the best global startups.
Since RITA’s inception, we have already carried out 2 scouting exercises, having reviewed more than 500 startup offerings from 19 countries. Our partners selected some of their products for further fine-tuning and verifying the proof of concept before launching pilot projects.
Since he joined X5 Retail Group as Deputy Director of M&A in September 2006, Anton Mironenkov has held various positions with the company until he was appointed to lead X5’s new Big Data department in 2018. This department is now part of X5 Technologies, a new business unit created in September 2020, also led by Mr Mironenkov with a team of more than 2500 professionals. Before joining X5, Mr Mironenkov managed various projects at Alfa Group and held senior positions at Troika Dialog and PricewaterhouseCoopers. Mr Mironenkov graduated from Moscow State University with Honours in 2000, having earned his degree in Economics.
