2016 can be officially declared the year of the chatbot. For  the last few months the amount of attention from the media,  and from developers and venture capitalists, has turned chatbots from an ordinary customer service tool to a hyped  technology, tipped to replace apps and become the all-  encompassing digital interface  between customers and  brands. And while a lot of the latest chatbot developments  are geared towards services and sales, customer support is  inevitably a focal point for the chatbot revolution.

So, what’s new?

Text-based messengers, or virtual chat assistants, have been a core offering for Customer Management Services (CMS) providers, chat platform vendors and in-house customer service departments for so long that when we asked a large outsourcing provider if their third-party chat application provides automation, they were surprised that there was any other kind on the market. Companies like IKEA, Ford, French retailer La Redoute, multiple telecoms, banks and government organizations had automated customer assistants on their websites in the early noughties. Many of these have since been deactivated, having failed to deliver on even the simplest support tasks. Hence the ‘Speak to a human’ signs on many websites.

What is new now is a combination of technological advances and changing customer preferences. Better machine learning algorithms, analytics engines, and availability of open cloud development platforms like Slack have made human-machine communication more effective. Tools like IPsoft’s Amelia are way ahead of their predecessors in their cognitive learning ability, making automated chats a lot more efficient in a typical support function.

Also significant is the way customers have changed their behavior in a relatively short period of time. Unlike Asian markets such as China, North American, European and Latin American customers have only recently accepted chat as a viable substitute for a customer care call. More and more users find chat to be a quicker, less intrusive, flexible, and richer interface with companies, as well as a whole generation of users whose purchasing and support behavior is shaped by mobile phones.

In our latest CMS outsourcing market analysis, the consensus among the 22 vendors we researched is that chat is the fastest growing channel, both in terms of revenue and volume. Many of the providers such as HGS and TeleTech are looking further into text and have implementations of two-way communication over SMS, while others such as transcosmos are expanding their ecosystem of partners, clients and services over messenger LINE. While fundamental, and with long-lasting effects, this channel shift is more gradual than many analysts are predicting. Voice is likely to drop significantly, but will still be the largest channel by market share until 2020.      

What will chatbots change?

For a start, bots will become an essential instrument in the digital customer service toolbox, and as we recently pointed out in a blog on self-service, most CMS vendors are ready and actively investing in this capability. In the more immediate term, as Simon Frank from Webhelp points out, brands will try to deploy chatbots on social platforms such as Twitter. This automation of the first level of customer communication will undeniably continue, but the ‘first level’ part is key here. Many who predict the end of the human call center misunderstand the role of the contact center agent, relegating it to little more than an answering machine. However, if a messaging service can replace a company’s customer service agents today, then that brand is well behind its competition in terms of its ability to interact, support and sell online or over the phone.

As chat usage increases and simple inquiries are answered by machines, human agents will need to become a lot more specialized, with communication skills defined by the channel, and offer a wider range of services on a single interaction (as Amit Shankardass from Teleperformance explains). Most CMS vendors are embracing this development: e.g. Concentrix recruiting for multilingual social media experts in Ireland, and Firstsource creating large-scale chat teams. Human agents will have to answer implicit customer questions and serve simultaneously as brand advocates, sales people, and retention experts.

In the not so immediate future, natural language processing and speech analytics research led by the likes of Nuance, Google and Xerox PARC are aiming for virtual agents that will understand accents and converse harmoniously at a contextual level. Until then, we can expect the following developments:

• Companies commissioning and developing their own chatbots, as they currently do with apps
• A proliferation of start-ups in this space
• End-users beginning to utilize tools such as Facebook M to interact with customer service departments
• Messenger interfaces such as Siri/Cortana more closely rivalling Google as the first port of call for internet-based support.

A major market analysis report on Customer Management Services has recently been published as part of NelsonHall's CMS program. To find out more, contact Guy Saunders.