By next year, over half of midsized and large enterprises will be using chatbots, Gartner predicts. It’s easy to see why: Chatbots leverage advances in artificial intelligence (AI) and natural language processing (NLP), and benefit from the fact that roughly 40 percent of all customer questions are about the same things.
Those are two reasons why chatbots save money, without affecting the quality of a customer’s experience. In fact, years of research by Forrester and other firms show consumers increasingly prefer self-service over waiting in a queue for a live agent. Chatbots accommodate that preference across all major channels, including mobile apps, websites and call centers. Meanwhile, advances in NLP and machine learning (ML) enable chatbots to keep getting better at understanding customers, even when they use everyday terms rather than industry-specific language.
As a result, contact centers can have live agents focus on providing premium service to customers who prefer talking with a human or whose questions are too complex for an automated response.
A Broad, Deep, Often Bewildering Solution Selection
With dozens of AI vendor and implementation options, organizations can be overwhelmed by all of the choices. That can stall a chatbot project, drive up costs or result in systems that can’t scale or evolve. To avoid these problems, organizations increasingly look for partners to help them understand where and how to implement chatbots, and then select the solution that best matches their goals.
For example, an organization may want to use chatbots to serve its Facebook following. Its technology partner might recommend IBM’s AI portfolio, which has tools for creating Facebook chatbots. The recommendation might also include using Watson Studio and Watson Machine Learning, which enable chatbots to recommend products that are most likely to interest each customer.
A partner can help identify which processes can and should be automated and integrate the AI tools with the organization’s existing systems — a service chatbot vendors themselves rarely offer. For instance, the chatbot’s ML engine could start by studying the organization’s existing product collateral and call center scripts. Once it’s learned enough to start helping customers, the ML engine uses those experiences to understand all of the different ways customers can pose the same question.
The right partner can leverage its expertise in dealing with AI to identify long-term considerations. For example, a municipality that wants to use chatbots to improve communication with citizens should choose a solution that can support future goals as well. For instance, the city’s chatbots may eventually learn to recognize when a person is in peril and then automatically route the inquiry to first responders.
That’s one more example of how chatbots are here to help — and here to stay.