June 29, 2017
Over the past year, Forrester clients have been brimming with questions about chatbots and their role in customer service. In fact, in that time, more than half of the client inquiries I have received have touched on chatbots, artificial intelligence, natural language understanding, machine learning, and conversational self-service. Many of those inquiries were of the “how does this stuff work” variety. But many other brands have seen very simple, rules-based chatbots and were wondering whether these technologies could really have an impact on high volume customer service environments.
The delta between those simple-to-build, but anemic in functionality, chatbots and conversational self-service tools designed for real enterprise environments is great. To truly meet a need for enterprise-grade customer service, chatbots must be able to understand what a customer speaks or types, pick out the customer’s actual intent, respond in a conversational manner and maintain the conversational state across multiple back and forths, and act on the customer’s behalf. These chatbot interactions must take place in a secure environment.
Additionally, the chatbot must have a way to seamlessly hand off the interaction to a live agent when it either cannot resolve the customer’s request or when it would frustrate the customer to continue the self-service interaction. Finally, as customers choose to have interactions across multiple channels (phone, chat, web, mobile, messaging, social, etc.), enterprises should be able to deploy their chatbots in multiple channels. Although the idea of truly ‘write once, deploy many” chatbots remains elusive, directionally, this is where enterprise chatbots need to head.
The vendors that offer such enterprise tools range from stalwarts that have been working with virtual agents for more than a decade to startups that have only existed for a little more than a year. To help our clients evaluate this emerging space, we winnowed the wheat from the chaff in the market and identified the 10 leading players in the market—the companies truly capable of providing the firepower needed for a genuine enterprise deployment. Then, in our new research The Top 10 Chatbots For Enterprise Customer Service, we evaluated those vendors in 10 critical criteria.
There are is one key takeaway from our research that is not directly related to the evaluation of vendor capabilities:
- Many of the enterprise-grade chatbot vendors mask the complexity of the technology from their clients. The care and feeding of machine learning and natural language understanding tools require expertise and several of the vendors take on this responsibility themselves. They employ the data scientists (and, in some cases, speech scientists) necessary to continually tune the systems and they provide that tuning as part of the service. For some enterprises, this will be a huge boon; for others, this black box approach offloads too much control of a critical function to a third-party. By contrast, other vendors provide brands much more granular control of managing the learning and continuous improvement process. You’ll need to evaluate your own appetite for tackling this work and make that part of your vendor evaluation process.
Check out this new research to help find which of these 10 solid offers makes the best fit for your business needs.
As a side note: this report marks the launch of a new research type and methodology for Forrester. Be on the lookout for more such rankings of technology vendors and service providers in emerging areas in the coming months.