sign with words "what's next" on it

How smart are bots now? Short answer: pretty smart. Consider how they are being used in the insurance industry. Seth Rachlin, executive vice president of insurance at Capgemini, reported that the industry is deploying bots on its most important and costly customer touchpoint, which is claims. 

Indeed, bots’ laundry list of accomplishments seem well able to handle the complexity of a claim process. "Bots today are already enabling the intake, adjudication, and settlement of simpler claims without human intervention and at incredible speed," Rachlin said. "Bots can analyze pictures of automobile damage, calculate repair estimates, and arrange customer appointments at 'insurer-authorized' service centers. Bots can read claim forms and associated medical bills, validate procedures performed against coverage terms, and reimburse members in a matter of minutes."

They can even interact with other forms of advanced technology, Rachlin continued. For example, after a hailstorm, bots can order roof inspections on damaged homes which are performed by drones, interpret the data they send back, and dispatch roofers to make repairs. 

The Bot Intelligence Spectrum

As bots continue to develop, expect to see even more advanced functions come to market. Most of us have become accustomed to the basic capabilities that bots offer, a level that Kevin J. Smith, senior vice president for Ivanti, calls intelligence level 1. Chat, for instance, falls in this category. "This is a good application for the capabilities of bots today because most customer service and support needs are basic questions that can be quickly dispatched by a capable bot," he said.

We are beginning to see the rise of bots that are capable of intelligence level 2, Smith said, in which a bot is able to go beyond answering basic questions. These bots can evaluate personal information and history and then take this information and make helpful recommendations to an individual, he explained. "We are beginning to see this capability on retail websites and when this interaction occurs, many of us will see that the recommendations are very good indeed. Companies like Amazon are careful about disclosing exactly how bots are used but is expected to be a pioneer in this area as a good experience creates a happy customer and creates potential revenue for the business." 

Smith expects to see bots on the level 3 intelligence spectrum in the next 2-3 years, bringing proactive and learning abilities to the bot interaction. He gives an example of a level 3 bot: this bot is able to contact a customer and offer a number of services based on an anticipated need as derived from data, trends and lifestyle. "Say a customer is passionate about coffee, a particular type of Guatemalan coffee, and a bot identifies the upcoming release of a new Guatemalan gourmet coffee, and notifies the customer of the release along with an overview of the new product and answers to questions the customer is likely to ask about the coffee and the company. As a coffee connoisseur, the customer will want to know where exactly the coffee is being grown, the flavor characteristics of the beans, how the coffee is being produced and what the annual yield of the farm is expected to be. The bot is ready with all of this information and more."

The bot, in short, has evolved from an inert technology, to one of customer adviser, he said. “We will find these interactions and recommendations truly remarkable and wonder how a bot could possibly know so much. Soon however, we will accept these increasingly intelligent, productive and even fun interactions with bots to be more common, and we should expect bots to display a sense of humor and an ever deeper knowledge of our preferences and habits as well as our passions."

In intelligence level 4, bots will have evolved into something akin to a personal concierge that answers questions, provides information, takes actions, runs errands, communicates with family, friends and colleagues, performs complex problem solving and makes recommendations among other possibilities, he said. 

Related Article: Top 26 ChatBot Builders for 2019

Levels 1 and 2 Not to be Overshadowed

None of these novel uses for bots at the 3-4 side of the spectrum should overshadow the uses that bots perform now, though, Smith added. “We need to remember that even answering basic questions is very helpful, in particular in cases where the volumes of interactions are high and the ability of a bot to offload some of this communications from our people is valuable in itself.”

For instance, Brian Westfall, senior HR and talent management analyst with Capterra, believes that chatbots should be seen as an integral staff member on HR teams. “Chatbots offer an immediate and impactful way to streamline an organization's HR efforts by streamlining simple, repetitive functions and freeing up employees to do more complicated or personalized work,” he said.

Improvements in Chatbots

For example, Westfall continued, organizations can rely on chatbot technology to answer questions from candidates regarding an open position general, saving recruiters valuable time, and giving applicants a better candidate experience. 

"Not only are these chatbots getting better at handling more complex questions from candidates — such as 'What's the company's work from home policy?' — but they're also getting better at handling shifts in the conversation,” he said. "If a potential applicant suddenly breaks off from a line of questioning about skills requirements to ask about salary, for example, it's not going to break the bot, and the bot can deliver that answer."

Related Article: How Much Does It Cost To Build an Enterprise Chatbot?

Enhanced Bots 

This spectrum of bot intelligence, though, is not quite so black and white with some bots in the lower levels of the spectrum being augmented with advanced technologies including artificial intelligence (AI). Coleman Parks, director of sales for Bovsi Studios, told us about a bot the company developed that uses Tensorflow scripts to automate and enhance LinkedIn business processes. "While other Linkedin sales tools allow you to create templates for messages, our bot actually learns from your messages and responses in order to create templates that will resonate more strongly with the recipient."

"Let's say your ideal client is a CEO between the ages of 35 and 55 in the IT staffing industry living in Florida whose company has less than 25 people. You’d then use our bot to connect with everyone who fits that demographic. As messages are sent, the bot would gather info on what type of response you get in order to learn what message format receives the highest conversion rate." Train the bot enough, Parks said, and it will even start sending messages only during the time of day your recipient is most likely to be in front of his computer. 

AI conversational voice bots are the latest bot technology businesses are using in their sales and marketing stack, allowing them to increase reach and touchpoints with prospective customers, said Ricardo Garcia-Amaya, founder of VOIQ.

"Conversational AI voice bots are helping businesses have more personalized, time-sensitive conversations with every prospect by automating sales and marketing calls. Currently, one of the largest catering marketplaces in the United States uses conversational voice bots to call and ask thousands of restaurants ‘do you cater?’ The food industry is one of many that tend to be more responsive over the phone vs. other channels like email or social media. With one click, this catering giant is able to reach thousands of their suppliers and soft-qualify them."