The chatbot — an insurer's customer-service ally

Insurers are increasingly deploying chatbots as part of their customer service offensive. Bots can help right across the insurance journey, but their success depends on seamless integration with other systems and how well they’re set up to learn to become human.

This is how the insurance world looks right now:

• Digitally savvy customers want to buy policies easily and settle claims quicker
• Insurers are looking for tag-team partners to improve their service
• Carriers with stretched resources are struggling to handle these growing demands
• There are multiple touchpoints with customers, but the data captured isn’t robust enough to enhance future customer interactions

But chatbots — computer software programs that can interact with humans using artificial intelligence — can help with all of these challenges. Their use is exploding across the industry, with Allstate, Geico, Allianz, and Liberty Mutual already putting them through their paces. Figure 1 shows their projected stellar growth over the next 5 years.

Digital natives are constantly searching for ways to make their lives more convenient, with little patience for anything that seems technologically obsolete. In the insurance space, customers — no matter where they are in the underwriting or claims journey — tend to have high customer service expectations. They want relevant, engaging, and contextualized interfaces. Texting and live chat has rendered unimaginable the frustrations of multiple interactive voice response routing prompts to answer simple questions.

Chatbots bridge the gap between customer-service demands and insurers’ customer-service abilities. It’s a win-win for both parties:

1. Conversational customer service. Chatbots can deal with written as well as verbal communications, with voice-recognition capabilities that enable human-like conversations. Combining these with human-like greetings and personalized responses based on existing customer information systems holds exciting opportunities
2. Enhanced engagement. Insurance customers are spoiled by the number of choices available to them. Insurers, in turn, need to find new ways to engage with them to reduce churn. Cleverly deployed chatbots can increase engagement, with focused responses to specific questions. They can also include plain text, images, internet data, web links, and interactive action buttons during conversations. Overall retention goes up as customers receive easily understood answers to simple queries
3. Improved insurer efficiency. Chatbots can be deployed more easily than mobile apps that need to be downloaded and installed. They are easily created using templates which allow drag and drop functionalities, have the ability to add plugins, and integrate easily with other applications to hit the ground running.

Insurers are looking at where and when they can bring chatbots into play.

Buying insurance, for one, can be made a lot less painful with chatbots outlining the steps involved, giving an overview of the options available, and helping with the final purchase. Chatbots can also reduce the confusion around different types of policies and cut down on jargon by highlighting differences between coverages, and recommending ideal coverage based on similar user demographics or state-mandated regulations.

By asking a few simple queries and connecting user needs with information available from other sources (eg., a social profile), a chatbot can also gauge patterns and present recommendations or information in an intuitive fashion, solving queries and creating a lasting positive impression.

They can also add value across the entire claims value chain. Logging a first notice of loss (FNOL), registering a request for repairs, scheduling an appointment for inventory inspections, or even giving an update on a claim’s status — these are all tasks that chatbots can handle. And field claims adjustors can use them when working on low-complexity claims, enabling query resolution in real time and speeding up loss assessment.

This makes data gathering easy. During the FNOL, for instance, chatbots can log the date, location, description, and even images of the damage. By using chatbots for customer interactions, insurers can reap the efficiency rewards — especially outside of office hours. Tag-teaming customer service employees with virtual assistants helps manage multiple concurrent conversations and ensures that customers get a quicker response.

Combining natural language processing capabilities with deep learning algorithms, these processes can be made much more intuitive, leading to the creation of highly human chatbots.

As with all new digital technologies, there are some teething problems.

• Language nuances. Mimicking human speech is a complex task. Mood, writing style, vocabulary, slang, and other factors all test chatbot abilities
• Victims of comparison. Humans tend to want better and better functionalities when dealing with technology and a halo effect can set in when comparing chatbot capabilities. Bots need time to gather data and properly analyze it before deciding if their performance is up to scratch
• Human preferences. As with human conversations, customers have come to expect a chatbot to be multidimensional, offering responses in the form of images, internet links, options, and even call-to action buttons. But not everyone likes this approach, and some will be deterred from further engagement
• Integration. A chatbot is intended to be able to work independently and without assistance. Full integration with other systems, plus access to insured, policy, and claim details is necessary to synthesize responses which provide meaningful information. Without investment in integration, customer service won’t shin

Chatbots are still in their infancy in insurance, but they’ve already made a big impression on some major industry players.

• AllState has the Allstate Business Insurance Expert, which targets its agents for commercial insurance selling processes.
• Geico launched Kate early last year to help customers, via text or voice-recognition queries, handle payment dates and other information.
• Allie, Allianz’s virtual assistant tool, is available online 24/7 to answer questions on its products.
• Liberty Mutual is the first insurance carrier to launch an insurance-focused service for Amazon Alexa that helps walk customers through the insurance process

Chatbots are just one part of insurers’ mission to deliver omni-channel customer service. A seamless user experience is still some way off, but the future looks bright as insurers start to embrace the new technologies available.

For chatbots to be a success, a holistic approach is key. The AI backbone that trains the engine to become more human over time will be the difference between happy or unhappy customers.

This point of view was authored by Girish Malik, vice president in the insurance vertical at Genpact.