E.ON has extensive experience with working purposefully with cutting-edge technology to streamline and improve the quality of its customer service whether by phone or in digital channels. One of the latest innovative inventions is a specialized chatbot that re-uses IVR logic to help customers asking for a deferment.
The specialized chatbot takes over when the “regular” chatbot has identified that the customer wants to postpone a payment and guides through the process directly in the chat window. However, the handover from one chatbot to another, and the smart tech under the hood, is not something that customers ever notice. They simply get their case resolved quickly and smoothly.
One of the key features of the chatbot is that it uses the same integrations as the corresponding IVR self-service. Therefore, the people behind E.ON’s customer service have been able to reuse the already built logic in a new channel. The project thus supports the stated ambition to resolve issues directly in the channel chosen by the customer, in this case the chat.
We spoke with Björn Tingelöf, product owner of ACE Knowledge at E.ON and one of those who worked hard to get the new function in place. With him, he has had a team of colleagues who have contributed expertise and experience from operations, development and testing.
The project is something of a pilot study for evaluating new technology. For that purpose the issue of deferral has been well suited as it is a relatively simple and well-defined service.
Recycling and modular thinking under the hood
To make this work, some key factors need to be in place. Good language comprehension with chatbot training on real customer issues to conduct a dialogue, authentication with BankID and integrations with other systems to do searches and perform backend operations.
The platform that E.ON has chosen to build on thus makes it possible to think modularly. The same logic and integrations can be used in different channels. Different bots, generalists and specialists can work together answering questions and offer self-service.
One flow for the customer
- From a customer perspective, the service can be described with a few simple steps.
- You surf to E.ON’s website and start the chat with the chatbot Kim.
- You explain that you want to defer a payment.
- The chatbot confirms your errand and asks if you have the opportunity to identify yourself with BankID (picture in Swedish below)
- If yes, then the deferral chatbot takes over imperceptibly, double checks that your social security number is in the customer register, checks that a deferral is possible, initiates BankID and offers to move forward the last payment date.
- If you approve, the chatbot also offers to send an SMS confirmation and you can enter your mobile number and it will be sent immediately.
- Here, the first chatbot takes over the dialogue and asks if there is anything else you want help with.
- Done!
The right skills, GDPR and statistics
When asked about challenges and lessons learned during the project, Björn answers that GDPR issues became a time-consuming part. Therefore, one recommendation to others who are considering similar solutions, is to discuss things such as data processor agreements early on. Here it was also important to ensure that personal data is anonymised correctly in the solution.
Statistics has also been a focus area of the project. It is vital to be able to follow up and verify that the solution really helps the customers as intended.
Another topic that Björn often returns to during the conversation is that having the right skills in the project has been a crucial success factor. Even if it is a lot of IT and technology, other parts of the business need to be involved. Not least, experience, knowledge and involvement from the work in customer service is essential to design dialogues and flows that the customer understands. “With a few thousand customer interactions behind you, you know what you do”, as Björn puts it.