(New) Workforce management: AI chatbot takes care of customers

the shortage of skilled workers in Austria is higher than ever before: every second company is suffering a loss of sales as a result. Customers react particularly sensitively to these service deficiencies. Artificial intelligence (AI) is helping companies in this tense situation: smart chatbots are now able to provide professional customer service. The wooden first attempts of the pioneering models are a thing of the past. The latest generation of AI chatbots independently classify inquiries, recognize customer concerns and needs, respond empathetically and relieve their human colleagues of time-consuming routine tasks.

QUESTION: Mr Urban, an AI chatbot puts itself in the customer's shoes in the call center - how is that possible?

Martin Urban: Artificial intelligence that responds empathetically to a customer inquiry is no longer a contradiction in terms. Two technological components contribute significantly to this. Firstly, a fundamental breakthrough has been achieved in the field of speech recognition. The latest generation of AI chatbots "understands" natural language utterances. Even convoluted sentences are no longer an obstacle - think of ChatGPT or Bard, for example. These so-called "generative pre-trained transformers" can be used very well in customer contact today. In concrete terms: if a caller asks about the status of their order, the AI in the system searches for a customized answer. The prefabricated response schemes of IVR menus are therefore a thing of the past. In addition to language understanding, the second important component is the programming of specific parameters that teach the AI empathy, so to speak. These parameters are tailored precisely to the company and the specific inquiries. Through speech recognition, the AI learns to distinguish whether a polite formulation or possibly an apology is appropriate. For example, if a customer asks indignantly where their order is, the appropriately programmed AI chatbot will first apologize for the wait and only then address the status of the order.

QUESTION: How do you know which questions the AI chatbot will receive in the call center?

Martin Urban: Before we start programming, we hold very intensive workshops with each of our customers. We analyze the customer journey as precisely as possible: What goals is the company pursuing? What products and services are available? Which target groups are addressed and how, and what expectations do the end customers have from the initial inquiry to after-sales support? Together, we work out in detail what the typical customer inquiries are and how the call center employees solve them. There are often call recordings in digital databases that lie unused. Artificial intelligence can examine such recordings in a short space of time using voice analytics. In this way, typical concerns and formulations can be identified. On this basis, we then recommend suitable entry points for automation to companies - the specific chatbot is programmed in close consultation with the customer. Important: We deliver everything from a single source and each of our solutions is tailor-made.

QUESTION: What would be typical entry points for automation?

Martin Urban: Some companies decide to use an AI chatbot at peak times, during lunch breaks or after hours. This approach is well suited to gradually introducing and optimizing the technology. In addition, the bots themselves learn and continue to improve customer interaction. Our recommendation is therefore to first automate sub-areas and then gradually integrate further phases and channels of the customer journey.

QUESTION: Is the goal to ultimately automate all areas so that humans become superfluous in the call center?

Martin Urban: No, quite the opposite. We always recommend a hybrid approach. It's about combining the talents of humans and machines in the best possible way. A self-service bot enables end customers to handle standard services such as appointment scheduling, reminders or shipment tracking independently and flexibly. However, there should always be the option of speaking to an employee for complicated queries. In this case, the call center agents benefit from the chatbot's preliminary work. This is because they receive the entire history as a basis of information - regardless of whether the inquiries were received by phone, text message, email or via the website. This means that the customer does not have to present their request again. The employees see the relevant customer information from all digital channels on their screen. The artificial intelligence also classifies the request. This additional information is a huge relief for employees.

QUESTION: Are there other ways to relieve call center employees through automation?

Martin Urban: In addition to the extra information for employees described above, there is a further significant reduction in workload because end customers are proactively satisfied and inquiries do not arise in the first place - according to the motto: "The best service is no service". With its customer experience management solution, KBC ensures that consumers receive all the necessary information automatically from the outset, for example on the status of their order or inquiry. In this way, the principle of "proactivity" frees up employees in the call center, who now have their backs free for troubleshooting and complicated inquiries. Human colleagues in customer contact are unbeatable for precisely such tasks.

QUESTION: In which industries have AI chatbots proven their worth in call centers?

Martin Urban: Basically, the use of AI chatbots makes sense everywhere in customer service, either to cushion the shortage of skilled workers or to save on repetitive tasks so that more time can be devoted to more appealing or complex inquiries. KBC sees the greatest potential in companies that have a lot of customer contact. For example, we worked with a bank to determine that a voicebot can free up the resources of around four employees right at the start of the conversation - with an average of 400,000 inquiries per year. That may not sound like a lot of relief, but you have to consider that this leads to an increase in quality (more time for consultations), offers more customer service (24/7) and also creates the basis for offering new services with the same number of people.

QUESTION: What is your conclusion?

Martin Urban: AI chatbots are an important tool for professional customer contact in view of the shortage of skilled workers. The breakthrough in digital voice processing has been achieved, making the technology very suitable for call centers. Many customer queries can already be resolved using self-service. In addition, proactive customer information through automated services further relieves the burden on employees. Smart automation solutions in customer service help to optimize processes, save costs and discover sales potential. KBC's digitalization experts deliver the entire solution tailored to each customer from a single source.