Much of the talk surrounding robots in the workforce has centered on robots performing manufacturing jobs or taking the place of cashiers. But the expanding capabilities for artificial intelligence (AI) mean that robots are spreading into other fields, such as sales and marketing. Chatbotswhose software enables them to converse with human counterpartshave been around since 1966, though only recently did they become part of consumers' lives. Early chatbots, such as AOL Instant Messenger's Smarter Child, were not particularly sophisticated, though, limiting both their ability to understand the nuances of human conversation and their appeal in practice.

As AI technology continues to advance, these limits are disappearing. A chatbot designed by Conversica effectively communicates with customers to generate sales. Companies that sign up with Conversica can pick the name and gender of their chatbot. Then, as sales lead come into the company, the dedicated chatbot gets in touch with the customer via e-mail or text message. If the customer shows interest, this chatbot puts them in contact with a human representative who can close the deal. The use of this tool for the initial contacts frees sales representatives from having to generate leads; they can devote all their time to interacting with customers who are actually interested in buying, instead of wasting time with unproductive cold-calling.

Since its founding in 2007, Conversica has grown to 140 employees who serve more than 1,100 customers in various industries, including automotive, education, finance, insurance, real estate, and technology sectors. Its corporate customers are famous names such as IBM, CenturyLink, and Fiat Chrysler of America. To date, its chatbots have sent approximately 250 million messages.

With this technology, Conversica also collects data that inform its constant efforts to make its software even more efficient. But the data do more than suggest ways to improve the chatbots' communication; they also reveal consumer preferences. For example, using female names from the 1980s and 1990s, such as Ashley and Stephanie, for the chatbots makes them more popular than those with other names. Jennifer is the hardest-working chatbot, according to how many leads it develops; Marie is the most engaging.

At a Connecticut auto dealer, a chatbot named Holly responds within four or five minutes to every potential lead. Approximately 60 percent of these contacted leads then respond, which prompts a connection with human sales associates. The dealer's ultimate goal is to engage customers in some sort of relationship with a sales associate, and Holly has helped it increase this form of customer engagement significantly. Furthermore, customers do not recognize that they have been talking to a chatbot; they often come to the dealership and ask for Holly specifically, though she is never in. Other businesses report that their chatbots have received thank-you cards or flowers.

The success and growing capabilities of chatbot software and AI suggest that Conversica will continue to expand its applications into customer support and service functions, as well as expanding the conversations to languages other than English.

1. Apply the Service Gaps Model to see how chatbots help provide customer service.