Data Has Always Been Pure Gold

Interview with Frank Brinkmann from Arvato Systems about Deep-Retail by Hannah Winter-Ulrich

New technologies such as Big Data, artificial intelligence, face recognition, and eye-tracking begin a new era in the retail industry. Based on the efficient analysis of data, they market the next stage of personalization - if cleverly combined. Frank Brinkmann, Vice President SAP Consulting at Arvato Systems, explains in an interview why it's time for deep retail.

Mr. Brinkmann, everyone is talking about data being the new gold. Why is that more to you than a stereotypical phrase?

Data is not just gold today. Data has also been the basis for successful business relationships in the past. Dealers have taken advantage of existing information about their customers. Just think of the innkeeper who knows the favorite drinks of his regular guests or the hotelier who knows which rooms his most loyal customers prefer. It's about more than excellent service. It's about giving customers the feeling of knowing them and their wishes - a high art. Those who can look after their customers in such an individual and personal way have the chance to increase sales. And nowadays, this is no longer possible without data.

At the moment, you get the feeling that the handling of data is inflationary. That begs the question: Which information is crucial? What do retailers need to know about their customers?

That's true, of course. Never before has there been as much data as it is today. However, this question is not easy to answer. After all, there are significant differences between retail companies and the level of digitization they have achieved. It generally makes sense to collect the following data: What gender does the customer have? How old is he? Is he a new or existing customer? What is his Geo-IP? Which device does he use to access an online shop? From which website does he get there? Which pages and categories are of particular interest to him? What is the monetary value of the shopping cart? By consolidating this information, it is possible to create significant customer profiles. In this way, retailers can address their customers with personalized offers and recommendations tailored to their wishes and ideas. Of course, one thing must always be guaranteed under all circumstances: Users must have agreed to the data collection and processing in advance. Otherwise, retailers will end up in the devil's kitchen.

Collecting data is one thing. Using them sensibly is another. How can retailers derive useful insights from their database?

Technologies, such as Big Data, artificial intelligence in general, and machine learning in particular, face recognition and eye tracking, are now highly sophisticated. Every single technology has the potential to create significant added value. But when retailers connect them, they can generate, analyze, and profitably use personal information in a new, compelling way. That is called deep retail.

Let’s look at the individual technologies separately. What role does Big Data play in an in-depth retail scenario?

Big Data is the supporting pillar. Whenever users inform themselves on the Internet, shop in an online shop, evaluate products and services - both on rating platforms and in blogs as well as in social media and shops - they leave behind digital traces. That is where Big Data comes in. Dealers are required to create their data confirmations from various sources according to their needs: Posts from social media, marketing surveys, inquiries to customer service, and the like. One example: When a new customer registers in an online shop, he enters his data in a form. In an SAP C/4HANA environment, this data is stored in the SAP Customer Data Cloud. When the customer later logs on to the shop via his social media log-in, social media profiles and inventory data can be linked, mandatory the user has agreed. All publicly accessible information from the social media profiles is related to the customer's click data in the retailer's portals, stored and analyzed. That creates a qualified user profile in the SAP Customer Data Cloud.

Retailers benefit from this concerning all customer-relevant processes, for example, in the service area: If customers send a request, the e-mail is automatically forwarded to the SAP Service Cloud, categorized and assigned to the relevant service team. Ideally, the application, supported by an AI-based algorithm, is even automatically processed in real-time and answered in the communication channel preferred by the customer.

What about AI?

With AI, retailers can go one step further and make data-based decisions. SAP Leonardo, for example, is an AI-based platform with which retailers can use machine learning and neural networks to derive precise insights from extensive data inventories and optimize their processes accordingly. For a system to learn from experience and continuously improve, an AI generates the necessary knowledge from unstructured data, such as comments and e-mails. To make predictions, it searches for recurring patterns. There are many practical examples: A machine learning system analyzes social media contributions. It recognizes that a customer wants a new coffee machine. The customer is then automatically presented with suitable products. Anomaly detection, which identifies deviations, is also practical. If a product suddenly sells particularly well in one region or target group, retailers can continue to fuel the trend or expand it into other areas or target groups.

Various studies have shown that AI has already arrived in the retail sector. The management consultancy Gartner predicted as early as 2017 that around 30 percent of global sales growth in digital retail would be attributable to AI by 2020 and PwC2 (unfortunately only available in German) concluded in 2018 that 44 percent of Germans expect more engaging shopping experiences in stationary retail through the use of AI.

A lot of data is generated by users uploading photos or giving ratings. Does this content also attract attention in deep retail?

Yes, this user-generated content is also included in the analysis. This is called sentiment analysis. It serves the goal of capturing the mood image in the target group. Because valuable insights can be derived from the mood and opinion of the users: Are products well-received? Which services do customers reject? The solution checks whether a statement is positive or negative. In this way, retailers can further develop their offerings in a targeted manner.

In the context of user-generated content, the smartphone plays a decisive role. Does it also do this in deep retail?

Here we are in the area of face recognition. Many smartphones are equipped with similar technologies - so far for device unlocking and authentication, for example, for Apple Pay. It is also conceivable, however, to use face recognition to record the mood of users to be able to offer them suitable products - which can be used to optimize the customer experience. Walmart, for example, has already registered a patent for a technology that recognizes the emotional state of buyers in the stationary market.

You also mentioned eye-tracking at the beginning. How does it work?

Eyetracking is another component that can be used to personalize the shopping experience. There are not only devices that can be mounted on standard screens, but also apps that access the smartphone's selfie camera and track users' eye movements. In the first step, retailers find out which areas of their shop are best suited to their needs, i.e., offers - looking at users, particularly intensively. In the second step, the system can then recommend those products that users are most likely interested in.

That all sounds very tempting. Are there any limits that retailers are not allowed to exceed, such as data protection?

Of course, there are such limits. Retailers must take their customers' concerns about data protection and security seriously. Recent scandals have significantly increased consumer awareness. Retailers and customers alike are caught between the desire for highly personalized shopping experiences and the legitimate interest in the protection of personal data. For this reason, retailers who seriously deal with deep retail must ensure absolute transparency concerning the handling of customer data. What data do they collect? How do they store it? For which analyses are the data used? How do they protect the information? If retailers act following the GDPR and maintain a serious approach to personal data, they can increase the trust of their customers and thus strengthen customer loyalty.

Hannah Winter-Ulrich (born 1982) is an experienced IT editor. She supports a wide range of companies in the B2B segment, including international corporations and large publishing houses, by creating individual communication solutions for their customers. In recent years, the studied Germanist, philosopher as well as media and communication scientist has acquired an extensive wealth of experience. Today, she passes on her extensive specialist knowledge to eager junior staff. As a mentor, she accompanies the training of prospective editors as part of a practice-oriented traineeship.

www.moeller-horcher.de