Exploring Data-Driven Retail with Metro AG

Learn what data-driven retail is and how to assess digital maturity in our guest blog from Pavel Ryukhov, Head of Global Store Operations at Metro AG

| Retail

Data Driven Retail


Retail is changing how it does business because of data and digital technologies. ‘Digital transformation’ is never far from the minds of retailers as they try to find ways to improve their business processes, operations and the experience of their customers. It can be a long road without a clear pathway.

Data plays a big part in retail transformation efforts and strategic decision-making. And data-driven retail has grabbed its fair share of column inches recently. So to help us understand this topic in more detail we asked Pavel Ryukhov, Head of Global Store Operations at Metro AG and a strong proponent and researcher of data-driven retailing, to share his perspective and experiences.

a photo of Pavel Ryukhov

What is data-driven retail?

To address this question, my colleague Alexander Shubin from myRetailStrategy and I, conducted research aimed at formulating a framework for data-driven approaches to retail management. This research not only provides insights into how to adapt strategies and tactics to the current state of a business but also demonstrates the process of constructing a suitable investment portfolio for transformation, ensuring genuine business value creation rather than expenditure on fashionable IT technologies.

The question “what is data-driven retail” is complex. Not from a theoretical standpoint, but rather from a practical standpoint.

There is a formal definition that tells us it is “a set of management methods and practices where decisions and retail processes are backed by data – with the ultimate objective of creating a remarkable value proposition for customers.”

It’s essential to note that no single company encompasses all aspects of data-driven retail. There are many different cases and success stories from the retail industry that we can reference as separate components of digital transformation. Innovations such as Pick&Go stores, smart pricing, computer vision for shelf management, augmented reality overlays in customer apps, machine learning for fraud detection, predictive maintenance for chilling equipment, and many others collectively form the ingredients of the recipe of the data-driven dish that we will cook all together over the next decade.


It must be said that business and technology aspects coexist together here. Technology acts as the enabler to accumulate and analyze data for use by the business to deliver measurable results in the areas of productivity, efficiency and experience.

We call this a ‘transformational data-driven strategy’. A comprehensive methodology to leverage data-driven technologies and management approaches that go beyond automating existing business processes. With the goal of creating a sustainable foundation for future changes and results, it unlocks the full potential of data-driven retailing.

What are the key areas that retailers need to focus on in order to enable a successful data-driven approach?

In our research, we identified eight groups of competencies that retailers should focus on.

  • Data-driven management – to demonstrate the progression of a retailer from operational, historical data towards longer-term qualitative scenarios and forecasts.
  • Data and systems – analysis of the data formats, data quality, data protection and ease of access to data within the company.
  • Organization, culture, and competence – to understand how the company organizes itself to manage the data preparation process at the department level, how that data is reflected in the culture of decision-making, as well as the employee level of digital competency.
  • The degree of (data) penetration into business processes – an examination of paper vs. digitized processes, and the proportion of system algorithms vs. humans within those processes.
  • Digital business models – to ascertain the degree to which a retailer is engaged in online business activities including marketplaces, e-commerce, subscription services, etc.
  • Customer experience and customer value proposition – to gain full insight into the digital interaction points a retailer has with its customers.
  • Investments and costs – to fully understand how the company handles its investments in data.
  • Innovation management – clarifies how a company builds an innovation funnel and how the selection, testing, and deployment decisions are taken.

Successful implementation of a data-driven retail approach is characterized by a significant share of revenue generated from non-transactional sales and substantial investment in data-driven technologies. Furthermore, it leads to a ‘culture of data’ where data and facts take precedence over expert opinions.


Digital transformation is a continuous activity, so how can retailers best evaluate their digital maturity and identify gaps in their processes?

That’s precisely the main topic of our research.

Retailers can use our methodology to find out at what level the company, process, or function is at. We defined phases ranging from data beginners to data champions. We assign stages based on the amount of data used in the decision-making at that point.

data-driven retail maturity model


In the first stage, managers rely more on ‘guts’ and intuition. In the last stage, the system presents the best options for tactical and strategic decisions. Every level is detailed with the appropriate criteria to make the position of the company or team clear.

I’ll give you an example.

Let’s take digital twins. Have you ever seen an entire store down to the last item reflected in a computer system?

Although we can already see some of the elements in ERP (Enterprise Resource Planning) and IDP (Inventory Forecasting and Demand Planning) systems, it still doesn’t exist. The progression from beginners to champions is not linear, some departments advance more quickly than others, and some take more effort.

Retailers today are still launching innovations that have been around for decades. News headlines cover projects such as electronic shelf labels introduction (the first commercial generation was brought to market in 1997), self-checkout (1992), and warehouse management systems (1975). This is perfectly normal. We should not forget that most of the market comprises well-established companies with history, legacy processes, and long-serving personnel. Integrating data in every facet of a retailer’s operations requires consistent, meticulous work.

Returning to our research, a critical milestone in achieving digital maturity is when the entire retail organization shifts its decision-making paradigm from relying on historical performance data to utilizing forecasts and scenarios.

Another crucial decision involves providing every working group and every project team with access to data and a full-rights member from the analysts’ team.

What are the benefits of data-driven retail?

In general, we would expect to see the benefits of data-driven retail in three key areas:

  • A direct impact on the top line when data-driven decisions help to sell more.
  • A direct impact on all cost lines including cost of goods sold (COGS), shrinkage, labor, maintenance costs, and administrative costs.
  • An enhanced customer experience, which over time makes a retailer more competitive.

In my current company (Metro AG), we improved the markdown of products near the end of their shelf-life using AI. Traditionally, the sales associate would locate the product on the shelf, run the best-before-date check process, choose the markdown price for the batch, print, and apply the markdown label.

Now, the decision is taken by the system based on the forecast of sales with the specific discount level – employees finish the task 25% faster now.

Managing pricing execution using smartphones

Besides a direct productivity impact, other valuable benefits are observed, such as the full visibility of the discounted stock in the system, the exclusion of the sold discounted units from the stock replenishment order to suppliers, the clear status of the ‘markdown’ frontline task on the dashboard of the store manager, and the prevention of potential fraud and price manipulations.

Other examples of the benefits that I’ve seen:

  • AI-powered predictive store equipment maintenance cut costs by up to 25%.
  • Inventory and demand planning system deployment reduced the stock of a retailer by up to 30% and consequently its working capital.

What key lessons have you learned from the adoption of a data-driven retail approach during your time at Metro AG?

Metro AG is not a typical grocer. The company operates in B2B markets in 21 countries, serving restaurants and independent traders. Metro performs most operational activities with wholesale stores and to some extent uses the store network to run a delivery business.

When we started our ‘digital operations’ project, we had in mind the idea of moving the store operations to mobile platforms. So, for all activities, our frontline employees would use mobile devices to get the same functionality, or better, compared to desktop computers. The core of the project included a new store inventory, messenger and task manager application, and the adjustment of the IT equipment infrastructure.

a grocery employee tackling store operations tasks with a smart device

The store inventory tool focuses on the fast execution of processes such as article maintenance, ordering, receiving, stocktaking, write-off, best-before-date management, and markdown application.

For the first use cases, I insisted that we provide our employees with an accurate reflection of their activities in the reporting app, ideally on the fly. When we adjusted our operations ‘authorization principle’ moving from ‘four eyes’ to ‘post-control’, that completed the picture and led us to entirely paperless operations.

It was not an easy journey. I managed a team composed of highly skilled employees from different functions, across multiple international locations and who were only able to meet online. And it was hard work aligning the new requirements to the existing IT infrastructure.

However, I’m glad to see that the efforts have paid off. Our frontline employees’ productivity has improved by an average of 50% across all realized store operations use cases. We have also achieved a higher quality level in those operations.

That’s not yet the final stage in our data-driven maturity matrix, but a huge step in the right direction.


It’s clear from Pavel’s answers that data-driven retail seems the direction of choice for the most forward-thinking and innovative retailers. The applications and benefits of using data for decision-making across retail operations can be wide-reaching. But knowing where to start and how to progress aren’t always obvious. Pavel’s framework offers a step-by-step process that retailers can follow to help them evaluate their current status on their way to becoming truly data-driven.

If you would like to know more details about Pavel’s maturity model for data-driven retail then you can access that here.

For more information on using Scandit as part of your data-driven retail strategy, you can find that here.