But first things first. What exactly is a chatbot? Wikipedia defines a chatbot as follows “a chatter robot is a type of conversational agent, a computer program designed to simulate an intelligent conversation with one or more human users via auditory or textual methods.”
Another definition I liked was one where a chatbot was defined as a replacement for every app you have ever downloaded. Today different apps answer different questions. An example would be a weather app. Yes, that app gives you a lot of detailed information about the weather, but the vast majority of users just want to know very basic information. Like is it going to rain? How hot will it be? And, this is where a chatbot excels. In the most basic terms, a chatbot is an application which you can ask questions using the conversational method.
So where does a chatbot live? Well, it lives inside what most would call a messenger app. Think AOL IM (AIM) from long ago or Facebook Messenger in the consumer space or Slack in the business world. Today we simply use these apps to communicate with friends or work colleagues. However imagine if instead of just being able to contact your human friends you could talk with a chatbot that could quickly answer questions in a conversational way. Perhaps even participating in the conversations with your human friends.
At its most basic the chatbot will directly answer generic questions. However, very quickly we see the lowly chatbot being able to access data specific to you and thus gaining what some like to call a memory or what I like to refer to as personalized contextual intelligence. Bottom line the more a system whether it be a chatbot or another kind of system knows about you the more useful it becomes. A perfect example of this would be the classic question will it rain. Well if the chatbot has access to your location information then it can answer a lot more intelligently than if it did not have access to your location.
Bottom line a lot of folks believe that given how quickly messenger products are replacing legacy communications methods like email and text that very soon chatbots embedded in these messenger apps will replace dedicated apps for search and customer service. This is exactly why Facebook and Google are so interested in chatbots. Given the massive revenue that comes from traditional technologies like search, the idea that something embedded inside a messenger app could replace all of that is a huge threat and justifies massive investment. The rumour is that Facebook will announce a chatbot store at its upcoming F8 conference and has been beta testing a bot inside its messenger client for a while now. The Dutch airline KLM has announced a customer service bot.
Now these are early days, and this was demonstrated by the disastrous events surrounding the Microsoft developed Tay bot where the internet quickly turned Tay into a racist. So as with any new publically facing technology, there are hurdles to overcome. That said the possibilities are infinite. Today's corporations have access to massive internal and external databases that hold information unique to all of us. When we call an 800 number today one of the first things a human agent will do is attempt to pull up our history in whatever form that takes to get the context surrounding our question. The objective is not only to make the call more efficient but to provide a more customized experience.
It is not a leap to imagine a world where chatbots with data hooks into our customer history and access via API connections to external data sources will soon be the first line of defence when consumers have a question. Because of the conversational nature of the interface chatbots align very well with the way that humans like to interact. Also the fact that these chatbots live inside the very interfaces where many consumers are now spending more and more time make the transition from traditional human agent to intelligent chatbot powered customer service even more inevitable.
So how do I see impacting the insurance industry? Well, the vast majority of inbound calls whether they are related to personal lines or small commercial level are to either make very basic changes to a policy or to ask a question which for the most part would be easy to answer so long as the agent or system has access to contextual data surrounding the caller. Examples of this would be policy endorsements like adding or removing a vehicle, requesting a basic certificate of insurance or asking a question related to current coverage. Questions like these could very quickly be answered by appropriately empowered chatbots that live right inside a messenger client such as the one powered by Facebook and others.
Now this is important not only because is a fundamental change to how insurance companies and brokers interact with their customers and prospects but it is also a significant shift away from where the interaction takes place. Today I see a lot of discussion in the industry around the need for customer portals associated with broker and carrier websites. My prediction is that these portals will become barren wastelands. Why? Because the way we interact with consumers is about to change in such a fundamental way that these static portals will act as an obstacle rather than a facilitator of interaction. As usual, consumers are proving, they will choose how and where they want to interact. With the advent of social media consumers have demonstrated that they have a voice and are willing to use it. Businesses have learnt that trying to silence customers only results in catastrophic blowback.
Now consumers are indicating that they wish to change the location of where they want to interact with businesses whether it be in the prospect stage or as customers. The idea that clients and prospects will somehow come to a company website or learn how to navigate some poorly designed generic portal is foolhardy. We live in the age of a socially empowered consumer, and they will continue to decide where and when they will engage. The age of the chatbot is upon us.