Denmark's largest energy and telecommunications group has combined human interaction with artificial intelligence in their customer service department. The initiative has resulted in reduced wait times for customers and more efficient workflows.
There is no doubt that customer satisfaction, in the telecommunications and energy industries, is closely linked to a well-oiled customer service engine. In fact, this is absolutely vital to staying competitive and maintaining a good reputation.
However, many people associate customer support with long call queues, frustrating transfers and lengthy FAQ pages that make them boil with frustration. And at first, an encounter with a robot might not sound like their dream solution.
Nevertheless, Norlys has chosen to use artificial intelligence to, among other things, identify a customer’s challenges more effectively so they are placed in contact with the right support officer who can help them resolve their problem.
Less time wasted. Happier customers.
When Norlys decided to supplement human interaction with artificial intelligence in their customer support, one of the motivations was the desire to accommodate customers more effectively. Senior Business Developer at Norlys, Martin Wickings, is in no doubt of the value AI provides their customer support:
– As a major energy and telecommunications company, we have experienced challenges, despite strong staffing and scaling power. Most recently, in connection with our merger into Norlys and move to Stofa, customers called in at the same time. Even though we had scaled up on employees, we still could not keep up, and customers did not get the service they should rightly expect. The chat feature would not only have mitigated this, but it also would have made a real difference. As we look ahead to more commonplace issues – such as unforeseen network outages – we will now be able to address problems through our artificial intelligence chat function, says Martin Wickings.
Norlys utilize artificial intelligence through a chatbot from SupWiz, developed in collaboration with NNIT. The chatbot can instantly deliver answers to customers and, in many cases, solve their problem. If the inquiry still requires the assistance of a human, the chatbot helps find the right employee:
– The way it works is that the chatbot asks the initial questions to identify the problem, after which it transfers the inquiry to the appropriate employee. This is an example of how we have partially automated our workflows, says Martin Wickings, and continues:
– In other situations, we have completely automated trivial tasks. For example, suppose a customer wants to know how to return their Wi-Fi router. The chatbot automatically delivers a return address and a description of how the equipment should be returned. At the same time, we ensure that working hours in customer support are spent on the more complex tasks. This results in less time wasted and happier customers.
Looking for more applications for AI
At Norlys, there is a great focus on improving the customer experience, and a strong belief that AI is part of the solution.
However, Norlys' customer service department is not alone in benefitting from AI. According to Martin Wickings, they are looking for AI applications to automate even more workflows in the group's many business units. And there is a strong reason for that:
– It is completely obvious for us to use AI in our work. Given our size, we have a lot of data that can be used continuously to test, train and evaluate the AI models. For example, when we started with the chatbot, we had several years of chat dialogue that we could use to train the language model.
However, the purpose of using artificial intelligence is not just to use top-trained AI models. It is very much about being able to provide an excellent customer experience while also staying ahead of competitors:
– Satisfied and happy customers may be the strongest competitive parameter, which is why understanding our customers' needs and desires is essential. With artificial intelligence as a tool, we can use our data far better, which will benefit our customers,” concludes Martin Wickings.
