A few days ago, Amazon introduced another smart speaker model. This new model makes Alexa, the voice assistant, available for use in cars. Speech recognition and voice-user interfaces continue to gain ground in our everyday lives. Insurers should no longer remain on the sidelines.
When Google released its new Google Duplex technology almost two years ago, it demonstrated what modern AI and speech recognition are capable of. This is because the system schedules appointments with customers over the phone without callers realizing that they are talking to a computer-generated voice. This is an impressive example of chatbot technology and how chatbots can communicate with customers automatically by voice or conventionally via keyboard – for example, in messenger programs, such as WhatsApp.As with many new technologies, the initial expectations for chatbots were greatly exaggerated. One of the forecasts was that a significant proportion of digital commerce would shift to purchasing via voice. But this shift never happened. However, the advantages of chatbots in customer communication are undisputed.
The benefits of chatbots in customer communication
Chatbots offer a wide range of advantages for direct contact with customers:
- Immediate responses to customers
- 24/7 availability
- Direct and rapid scaling
- Reduced costs
- Filtering out routine inquiries
- Automation of inquiries in social networks
Since chatbot systems can log user inquiries, customer inquiries reveal a lot about open questions, concerns and wishes. This data helps to improve chatbots, but it can also be used by product management to generate new ideas for products and services. For example, when there are frequent inquiries on a specific topic.
Contrary to what employees may fear, chatbots are usually not used in order to eliminate staff. Frequently, the focus is on reducing employee workload and expanding customer communication. Instead of having to deal with routine inquiries, employees have more time to answer more complex inquiries, while chatbots, for example, provide customers with information on pricing.
Many insurers are gaining experience with chatbots
A number of insurers have gained experience with chatbots as a method for offering a variety of solutions.
Geico is one of the largest car insurance companies in the United States. The company uses a virtual assistant called Kate to help its customers. The interactive system responds to voice and guides insurance customers through the services offered by Geico. In this way, it provides customers with specific information on available products and the corresponding contracts.
As a purely digital provider, Lemonade focuses on AI-driven customer service. Lemonade uses several chatbots at the same time. These chatbots can often complete a customer's application process within 90 seconds. However, the simple pricing structure also makes this easier.
Helvetia uses a chatbot to process claims via mobile devices. Helvetia's promise to its customers is that it will respond extremely quickly in the event of a claim. One of the system's challenges is user authentication. The chatbot must find the right mix of questions about the customer's contract data and personal information.
As a multinational company, Axa relies on a whole range of chatbots in different countries. These chatbots handle traditional routine inquiries, but they can also answer specific questions. In the context of insurance for legal expenses, for example, the systems answer general customer inquiries, such as questions relating to laws concerning the rights of neighbors.
AI is the foundation
Some companies now offer modular systems for chatbots. Customers can assemble their own systems with just a few mouse clicks. But this creates false expectations for companies, such as insurers, whose products require explanation.
The quality of a chatbot's answers depends on the quality of the AI behind the chatbot and the material used to train it. The chatbot has to accurately interpret customer inquiries, even if the input is not clear because consumers don't know the product name or, because they're in a hurry, they use expressions which are difficult to understand.
Chatbot projects in the insurance world are and will remain very individual. On the one hand, these chatbots need to be directed towards specific company goals and on the other hand, the AI knowledge databases need to be trained according to the company's material. While a chatbot is in the learning phase, there should be more than enough of this material available in every insurance company (customer inquiries and their answers, evaluations of typical conversations, information material, etc.)
The implementation of chatbots, regardless of the channel used, is an example of the need to break down data silos so that there are no barriers to AI learning.
