EOS Forum 21

EOS Forum 21 was already the 15th edition of the popular networking event for practitioners in receivables management. The 2021 topic: the use of artificial intelligence and data in receivables management.

• Which approaches and tools for using artificial intelligence (AI) in receivables management does the EOS Group work with?
• What has been the experience of putting them into practice?
• Which success factors must be considered in AI projects? 

In the keynote, Joachim Göller, Senior Vice President at the EOS Center of Analytics, gave insight into the methods and procedures with which his team in Hamburg identifies possible AI applications, develops them for concrete use and tests their effective use and added value “for real” in pilot projects. If an AI application proves itself in practice, it is offered to the individual country companies of the EOS Group for introduction. Examples are the rapid valuation of large real estate portfolios, the chatbots (Tom and Lea), best-next-action analyses for classic collection cases, and the scoring model for assigning cases to automated or manual workflows.

Interestingly, Göller named the “human” factor as one of the most important success factors in AI projects. Likewise, involving the internal data protection officer early on is worth its weight in gold, because their expertise also reveals opportunities for the anonymized and pseudonymized use of data – keyword privacy by design. Furthermore, the management method Objectives and Key Results (OKR) has proven very effective for new AI projects at the EOS Center of Analytics, because it prevents getting lost in the endless possibilities of the huge AI universe, with negative consequences for the cost-benefit ratio.

Pepe Kondakow, Director Operations of EOS Schweiz AG, used concrete examples of scorings, payment plan models for determining optimal installment agreements, and best next action to show how AI solutions are applied in practice in receivables management. With digitalization and the use of AI, EOS is pursuing, on the one hand, better accessibility and faster service, in future also thanks to self-service offerings for customers. On the other hand, AI applications help make individually optimal decisions and use resources as well as possible. This includes relieving collection specialists of ever-identical routine tasks so they can take care of more complex cases.

Even with the growing use of robotic process automation (RPA) and AI, EOS Schweiz continues to value careful control by a human before there is any interaction with defaulting payers: after all, sensitive information is always involved – and the reputation of the creditor client.  

The moderator Nina Havel then invited everyone to the panel discussion. The two speakers were joined by Wilko Wiesenhütter, Head of Credit Risk Management at Swisscom AG, and Melanie Blattert, Head of Receivables Purchasing at EOS Schweiz. Swisscom already uses AI applications in customer service today. According to Wilko Wiesenhütter, customers expect a quick solution to problems, even for more complex customer concerns. AI helps recognize the customer’s concern and offer solutions, even before the customer gets in touch. This creates a customer-specific experience that increases customer satisfaction and at the same time saves costs for Swisscom. However, at Swisscom too, concerns for whose resolution empathy is a key skill are always handled by people. Melanie Blattert followed up with the statement that AI systems are a helpful complement in receivables management, but that people will always remain important. Interestingly, Swisscom’s experience with its chatbots shows that customers with “simple” concerns such as a payment deferral or an installment payment accept the self-service model with a chatbot very well. But they are glad when they can discuss more complicated cases with a competent, empathetic employee. 

Blattert pointed out that AI can also help preventively to avoid accumulating new debts because of installment agreements. In Belgium, EOS already uses a solution with which a budget calculation can be created for defaulting payers. EOS Schweiz intends to make this budget calculation software available to customers in arrears on the new portal for defaulting payers, so that when they enter a repayment plan they can see immediately whether the desired installment plan is realistic for them at all.