The potential certainly seems rife. With more channels across which to expect support, the noise-to-signal ratio is rising and it’s increasingly difficult to pick out messages that merit a response. Robots might indeed have an advantage over their human counterparts. And as more customers seek self-help solutions, a fully automated, “machine” customer support seems possible.
But is it inevitable or likely?
The prevailing wisdom seems to suggest otherwise.
Machines excel at routine functions and can give customers instant, frictionless answers to common questions. This offloads traffic from customer service and speeds up resolutions. To this end, Dominos and 1-800 flowers have famously allowed customers to order through chat bots via Facebook Messenger. But what happens when these customers go off-script?
If Microsoft’s AI chatbot fiasco last year taught us anything, it’s that robots have their limitations. If people know that they’re dealing with robots, they treat them differently and often try to game the system. Drivers, for example, happily cut off risk-averse self-driving cars, and Microsoft had to quickly pull the plug on their chat bot after online trolls taught it to repeat racist comments.
Chat bots, no matter how intelligent, still need a human chaperone to inject commonsense. The risks are simply too high, especially when social media mishaps carry the potential for PR pyrotechnics. Human customer service agents simply don’t fall for programmatic manipulation and while support bots can be useful, but they’re not a complete replacement.
If proactive talking machines aren’t a complete answer, neither are passive self-service ones. Forrester reports that consumer usage of self-service support options like forums, communities, virtual agents, and IVR nearly doubled between 2012 and 2015, but this raises a question as to how much of that was driven by the brands themselves. During this self-service explosion, 52% of consumers still switched providers because of poor customer service, costing U.S. businesses $1.6 trillion, reports Accenture. This sounds like an effort issue, if you ask me.
Through their actions, consumers reveal a distaste for brands who conceal their support contact information behind self-service forums. While this practice at first seems to helpfully divert traffic away from expensive legacy channels like phone and email, in the end it merely pushes them to social media where their problems bloom into public crises.
Consumers want resolutions, not options, and when the passive, sit-back-and-wait approach isn’t enough. Consumers will still need someone to reach out.
The best answer would seem to draw from the best of all worlds: machine efficiency with good old fashioned human engagement. Nowhere is this on better display than in social customer care.
Forging genuine, human bonds with consumers is the greatest advantage that social media offers brands today. This is a critical advantage for all brands, as "emotion" is among the top three necessities for modern customer service organizations, reports Forrester. Customers simply don’t want to fall into the icy grip of unfeeling machine. Social media, moderated by humans, has thus turned the stale, inhuman aura of 1-800 phone trees or email into heartfelt, emoji-stuffed responses and genuine connections:
The empathy and context-specific apologies within this approach simply can’t be faked. Those consumers that receive positive human resolution on social media are 4x as likely to recommend your brand, and with machine automation behind it, those human agents can handle 4-8 x as many issues per hour.
The best solution, then, is neither man nor machine: it’s a bit of both. Trained, empathetic agents supported by intelligent automation can both cull through the fire-hose of consumer messages to handle the routine traffic effectively and identify those conversations that need a personal touch to provide some <3.
The future then isn’t man vs machine; it’s man (or woman) powered by machine.