Businesses around the word are applying intelligent automation and artificial intelligence to realize efficiencies and create competitive advantage. Insurers large and small have built chatbots or virtual assistants (VAs) to reduce overheads, improve customer service, and accelerate product deployment.

Recent Accenture research suggests that this is just the start. Chatbots and VAs have strong synergy with messenging apps like Telegram, Facebook Messenger, and WhatsApp. Such apps are now the most popular form of digital communication around the world. Messenger alone has more than 1.2 billion active users.

As we’ve seen in this series, chatbots and VAs are becoming more attractive to insurers of all kinds. These tools may eventually become indispensable. But any new technology demands a clear strategy. Here are five principles for insurers to keep in mind when getting started with chatbots and VAs.

1. Keep it simple

One of the biggest advantages chatbots and VAs offer is efficiency—for insurers and customers both. Interactions with these new tools need to be as efficient, short, and painless as possible. Systems that require users to create new accounts or sift through tutorials defeat their own ends. If engaging a chatbot or VA feels like a chore, users will look elsewhere.

2. Don’t get lost in the crowd

Chatbots are already in common use across the insurance industry. VAs are less widespread, but still far from rare. Both are likely to proliferate. Insurers need to consider what will make their chatbot stand out, how users will discover it, and what will make it essential compared to the competition.

3. Keep it consistent

Customers increasingly expect a seamless communication experience across all interactions with a business. To meet this expectation, a chatbot or VA needs to be integrated across messenging platforms as well as mobile and web services. These tools should “talk” to every channel in an insurer’s ecosystem.

4. Keep it secure

People won’t use services they don’t trust with their data—especially for insurers, who are often privy to very private information about their customers. Strong security needs to be front-of-mind for every insurance chatbot or VA.

5. Don’t misunderstand your customers

Chatbots and VAs are wonderful tools up to the point that the user wishes they were interacting with another human being. Seamless, simple AI interactions will be critical, and customers will quickly turn away from “smart machines” that don’t respond to straightforward language.

Yet technology is still a long way away from producing an AI that can independently handle the most complex customer interactions. Insurers need to build fallback procedures into their chatbot and VA systems that will quickly shift users to human support when called for.

These principles will help insurers unlock the tremendous power of chatbots and VAs. The benefits of these tools apply to three areas of business in particular:

  • Faster, cheaper customer interaction. Labor currently accounts for nearly two-thirds of the cost of running a call center. Insurers looking for efficiencies here can turn to chatbots and VAs. They have the potential to be almost twice as fast as phone-based support, and cost much less.
  • Automating operations for efficiency gains. The most obvious application of chatbots and VAs is customer service, but they have many more uses. From screening job applicants to filling out timesheets, they can help insurers improve operations in many facets of the enterprise.
  • Making sales and marketing teams more effective. Like other uses of AI, chatbots and VAs can free up the workforce to concentrate on high-value tasks. Tools like Allstate’s ABIe allow insurance agents and marketers to ditch drudgery and focus on creating value.

This concludes our series on chatbots and VAs in the insurance industry. More general chatbot information is available from Accenture here.

One response:

  1. Your article has neatly explained everything Erik, I can’t agree more than this. Train a bot to become independent could be the biggest challenge that any development team experience. We are working on a writing bot targeted to Integrate into the digital journalism ecosystem and respond to all sort of writing patterns, however the biggest challenge exists when the bot has to write content on recent happenings and hot topics, too an extent it’s addressing the partial objective, but not responding fully functional when it asked to perform on critical aspects. Seeing the behavior and pattern, I agree completely that they are not 100 percent independent yet. At some point human intervention is required to accomplish an episode for customer engagement.

Submit a Comment

Your email address will not be published. Required fields are marked *