We use AI in an experimental format

Dmitry Marykin, Director of Customer Service and Digital Experience at the M.Video-Eldorado Retail Group told about e-commerce, retail and digital technologies for large and medium-sized businesses as part of the podcast “Practice Days”. We share the most interesting points of the conversation.

— Dmitry, hello. Thanks for stopping by. How is the company doing today? Any up-to-date numbers? I found information online for last year and didn’t go into detail about the last quarters. What can you share?

– Hello. We are an open company, we publish our reports, so the figures can be found in open sources.

Almost every resident of Russia knows M.Video. We have 54 regional warehouses, 8 central distribution points, more than 1,240 stores across the country – that is, we are a large-scale federal network with a large infrastructure. To understand the scale, the revenue of the M.Video-Eldorado Group for 2023 amounted to more than 434 billion rubles, and total online sales amounted to 71% GMVincreasing by 18% to 386 billion rubles.

There are already almost a thousand merchants on our marketplace, the average bill is growing, we feel good and confident.

What's interesting is that every retail store here is still DarkStore for fast delivery of goods within 1-3 hours – by the way, we are the country's leading experts in the delivery of large-sized equipment. We continue to develop OneRetailour hybrid business model.

In large cities – Moscow, St. Petersburg, Novosibirsk and Yekaterinburg, as part of the development of our service, we launched a branded zone M.Master for repair of household appliances and electronics in stores. We continue to expand the brands and equipment that we repair: Apple, Samsung, Huawei, Realme, Xiaomi, and others.

— Your title sounds like Director of Customer Service and Digital Experience for a retail chain. Can you tell us more about what you actually do with such a complicated title?

— In general, my position can be divided into large blocks by departments:

1. Contact center with a customer service unit inside it. It often happens in companies that customer service as a team is led by the contact center team.

2. Customer experience management, which includes research, development of a feedback platform, collaboration with functions, including with colleagues who invented and are implementing M.Master.

3. Developing the experience of a retail employee and, in principle, every employee who interacts with our client, as a response to the OneRetail strategy.

I promote the understanding within the company that every employee, without exception, is responsible for the customer experience. It all starts with a culture of values; continues with who and how we hire; how we further train and develop the employee; what tools he works with; what promotions we launch; how

our stores look like. There is not a single function in the company that is not responsible for and does not influence the customer experience.

— Tell me, please, what is the most important thing for customer service from your point of view? What points would you highlight?

— Firstly, this is the so-called optimal customer path with a seamless transition. This is when the customer makes a purchase in a way that is convenient for him and when he needs it, and not when the company forces the customer to “finish the shopping”.

Secondly, it is important for the client to be sure that the goods he sees in the online/offline showcases are expertly selected, available and of high quality. And if suddenly something goes wrong, the retailer will resolve any issue that the client may have as quickly, efficiently and in a convenient format as possible.

Thirdly, it is the fulfillment of promises. And here we have implemented a metric throughout the company perfect order — i.e. an ideal order, where we check absolutely every delivered order for compliance with the promises made to the client.

Fourthly, it is feedback. We are currently conducting both large-scale studies, such as a large customer journey map, and tactical and operational surveys, from which we use the information every day in our work. These are assessments NPS, CSImonitoring of requests: all this gets into the feedback platform, is sorted out, we see at what stage of the customer journey something has changed, what we need to improve. Then we immediately interact with the functions that could be the conductors of these changes. And then we observe the dynamics.

— Please tell me, based on all this, what parameters, what numbers are key for you?

— In fact, I don’t single out any individual figures. Because for me, the figures complement each other. That’s why we look at the complexity, which includes NPS, CSI, and the share of requests from receipts.

— Tell me, please, do you evaluate the results of each project in terms of profit? Is there such a metric in your dashboards? If so, which of the completed projects brought the company the most income? I understand that the client is more important, but if we consider the result precisely in terms of profit for the company.

— We, like any company, fully understand the importance of customer service, but we cannot always accurately measure its value in monetary terms. Money is a consequence of good customer experience and service. We evaluate the effectiveness of our projects primarily by their impact on key metrics. Why is that? Because we have found a correlation, a mathematical proof, confirmed by a year and a half of research. This metric is an axiom for us, so we are working on improving it.

— What project are you most proud of?

— All our projects, as a parent is proud of all his children. Last year we implemented 147 initiatives. Perhaps this is a lot, but they were of different scale and size. Of these 147 initiatives, 97 have been fully completed at the moment. In this podcast, I told you about three of these 97. It is difficult to single out one project.

— Does your team work with artificial intelligence? What are the prospects for such projects? I am sure that in 5-7 years this will become a standard. What stage are you at now?

— We are using AI in an experimental format. We wanted to find out whether GPT and other large language models can help us in consulting clients and whether they can be trusted with data about our products. Will they be able, for example, to select the right multicooker for a client? For example, each employee has a mobile sales app, which we are proud of, and it is recognized as a successful practice in the market. We have more than 200 thousand SKUs, which makes working with them difficult. To help our employees quickly provide accurate and expert advice, we started an experiment with AI.

— What analytics tools do you use? Are these ready-made solutions or something based on your own database?

— In our feedback collection and management system, we actively use ML models, which, as far as I know, are self-written. We pay a lot of attention to their accuracy. The level of recognition of customer feedback is now more than 93-94%. This allows us to build a process based on feedback analytics and ML models.

— If we summarize everything you said, then your main job at M.Video is monitoring feedback collection, clustering by requests, ranking and segmentation, monitoring initiatives and interacting with different departments. Is that right?

— You've only described 30% of my work. The strategy for managing customer experience and service is a systemic action plan. We not only coordinate all functions related to customer experience, but we are also a project and product office,

which generates and implements new solutions. The remaining 70% of the work is related to the contact center. We also actively cooperate with Data Office and have achieved a high level of automation with a CSI of around 95%. This allows employees to focus on the customer and the product, simplifying processes.

— What advice could you give to a medium-sized business engaged in retail?

— It all starts with culture, mission, and values. The mid-sized business needs to make sure that the customer is an important part of the equation. Working with customer experience strengthens the business, increases sales and market share. It’s not a sprint, but a marathon without a finish line — a constant process with delayed results. Second, avoid analysis paralysis, move to practical use of data as quickly as possible. Third, know how to negotiate and complete processes.