INTERVIEW: Advanced data analysis as a source of success in the retail industry

An interview with Kamil Folkert, CSO at 3Soft and a co-founder of Occubee was published on Computerworld.pl and in the paper version of the Computerworld TOP200 report, 2022 edition.

In the interview Kamil Folkert talks about the benefits of implementing a data-driven approach in business and the best way to start the adventure with artificial intelligence in the retail industry.

How can the retail industry benefit from the potential of data?

Advanced data analysis based on artificial intelligence and machine learning is becoming a shield for retailers in the face of new challenges – economic and geopolitical turbulence, effects of the COVID-19 pandemic, broken supply chains, shortages of raw materials and human resources. At the same time, it allows them to build a competitive advantage and to cope more effectively with new market conditions, dictated, among others, by dynamically developing e-commerce and growing consumer demands.

Working with clients representing different sectors – FMCG, home & garden, fashion – shows us that many of them are already implementing a digital transformation process, e.g. by using business intelligence systems, but there are also those who are still struggling with data silos in Excel and identify a strong need to implement a strategy to change their approach to the data they have. They want to improve processes and evolve into a data-driven business, preferably in an evolutionary, not revolutionary, way.

Retailers see the real value in data, and they view big data and artificial intelligence as game changers, the ticket to effectively scaling the business, reducing costs and increasing profits.

However, the path to achieving these goals varies, as each company faces slightly different challenges and formulates its strategic and operational goals differently. Our work with retailers therefore begins with identifying business opportunities and determining how, using mathematical modeling and advanced AI algorithms, we can improve process efficiency.

Is artificial intelligence reserved only for top players?

It’s not just corporations and international retailers that can benefit from the power of artificial intelligence. Our Occubee platform is the best proof of that. We have developed it primarily for medium and growing retailers who have historical data that allow predicting future business events. In this way, we lower the barrier to entry and make it easier to implement AI in smaller scale businesses.

Analyzing data with Occubee does not require investment in additional point-of-sale equipment or expansion of your own data center. The SaaS model makes this approach financially attractive and business efficient. Operational risks related to the continuity of existing IT infrastructure and business processes that have been optimized over the past years are minimized. Retailers can therefore achieve measurable business benefits at a relatively low cost, shortly after implementation – not by revolutionizing the way they operate, but by enriching existing processes with the knowledge of future business events that resides in the data.

Artificial intelligence is no longer an extravagance or a futuristic curiosity, but the next stage of development and even a necessity. This applies not only to international retail chains, but also to retailers representing the mid-market.

How does being data-driven affect business?

By analyzing billions of receipt data, information about thousands of products in different locations, and dozens of retailer-specific factors, we are able to predict customers’ future purchasing decisions. This is done, of course, with a certain probability, just enough to make sound business decisions based on facts rather than intuition.

The use of machine learning models makes it possible to generate hundreds of sales forecasts per day, taking into account many internal (e.g. price lists, store locations, marketing actions) and external (e.g. weather forecasts, public holidays, special events) variables, which, when compared with individual model training at the level of product/store combinations, far exceeds human capabilities. In addition, the use of the latest technologies in the area of cloud data management ensures effective monitoring of data flow in real time. This makes it possible, among other things, to generate alerts in case business anomalies are observed in the data flowing in on a regular basis.

Retailers with advanced AI algorithms can reduce overstocks, allowing them to free up frozen capital. At the same time, they can increase product availability by reducing out-of-stocks, thus minimizing the risk of lost sales.

It is also worth noting that by improving business results, we support retailers’ actions in the field of sustainable development. For example: better planned logistics on the warehouse-store line and optimization of loading space of delivery vehicles result in reduced carbon footprint. High-quality demand forecasts in the FMCG sector contribute to reduction in waste, which is particularly important e.g. in the case of food with a short shelf life. In the fashion industry, more accurate forecasts equal using less water and energy to produce clothes.

Artificial Intelligence supports the ongoing revolution in sustainability. It is worthwhile for every retailer today to ask themselves: do I want to be a participant and beneficiary of it, or just a silent witness?

What does AI implementation mean from the perspective of retail chain employees?

Artificial intelligence cannot replace humans, if only because not all the regularities regarding the future of a given business are reflected in historical data. In such situations, the role of an expert and their perception ability is a key element in maintaining the operational efficiency of the business. As an example, let me use the analogy of an autopilot in an airplane. The fact that such a system exists has not caused airlines to abandon human pilots on board. The autopilot can be responsible for the bigger, though technically probably the easiest, part of the flight, allowing pilots to focus on those activities that require their intervention. The same is true in retail, especially when starting out with advanced AI algorithms. It’s worth using them to automate repetitive yet large-scale tasks, which often account for tens of percent of the activities necessary to ensure business continuity. This will free employees from routine tasks and reclaim time for things that require their unique attention, creativity and expertise.

Where should retailers who want to implement data-driven strategies start?

I recommend free workshops with 3Soft experts. During the meeting we learn about the specifics of a given business, analyze available data, identify needs and address concerns. Together we look for business opportunities and consider possible solutions. Thanks to that we discover areas that require optimization, which are of strategic importance in the context of increasing the company’s value.