Blog Viewer

The New World of Robots in Customer Service

By Tema Frank posted 11-02-2015 07:21 PM

  

Robots in Customer ServiceI recently read the book, Rise of the Robots: Technology and the Thread of a Jobless Future, by Martin Ford. Then I followed it up with an interview with Bent Dalager, Managing Director for Financial Services Nordic at Accenture in Copenhagen, who told me that robots are already replacing tens or hundreds of thousands of customer service related jobs, such as insurance claims handling, and are fast encroaching on the contact center.

If you already follow the Frank Reactions blog, you’ve seen my thoughts on this, but if not, here’s what I took away from our two-part interview. (If you’d rather listen than read, you can hear my interviews with him at http://FRitunes or at http://frankreactions.com/show)

Spy Mode Lets Robots Teach Themselves

For as little as $10,000/year you can get and train a virtual robot to handle masses of queries or back end paperwork, at much higher volumes, speed, accuracy and consistency than humans can manage. Global consulting firm Accenture has already replaced 8,000 full-time equivalent jobs with “virtual robots” and another 16,000 will go this year.

Technology Breakthroughs

The breakthroughs making this possible have been:

  • The ability for the robots to use self-learning, or cognitive computing,
  • A new-found ability for them to switch between computer programs, so you don’t need special software to connect all your systems. 

In the back office, for example, you no longer need IT to do complex integration programs, and you won’t have to ditch your expensive legacy systems. “Spy mode” lets the computers watch how your human agents process things and do the same, even when it means moving data across systems. (I must admit that somehow it does seem unfair that staff are having to train their automated replacements.)

Impact on Customer Service?

When it comes to customer service, robots can do first line triage of problems, solving many of them without having to escalate to a human being. And they learn from what happens if problems are escalated, so over time fewer and fewer queries need to get the human touch.

There’s an insurance company in the Netherlands, says Dalager, that already has 80% of the incoming customer service chats handled by a robot. This may sound dystopian, but look at the bright side for consumers: a robot can handle multiple chat queries way faster than a human.

But Don’t Customer Service Reps Need Empathy?

They do. This is the challenge preoccupying leading artificial intelligence companies.

According to Dalager, robotic technologies are getting much better at being able to recognize and respond appropriately to customer emotions. When you combine the self-learning skills of IBM’s Watson with the increasing empathetic and context-sensing abilities of IPSoft’s Amelia, the robotic revolution will go mainstream.

That’s only a year or two away, thinks Dalager. Amelia can already use context instead of just key words to be able to tell the difference between “I want to see red boots” and “I don’t want to see red boots”, and to detect the emotional pitch of a customer’s voice and adapt what she is saying accordingly.

By 2018, 30% of administrative jobs will have been replaced by robots, says Dalager. And the folks who couldn’t make it into med school may have the last laugh, because by 2023 about 1/3 of  doctors, lawyers and professors will be replaced by people with far less education, tapping into the cognitive abilities of virtual robot partners to supply the ‘brain power’.

If Robots Are Smarter & Faster, What Role Is Left For Humans?

Although researchers are making progress on machines that can empathize with customers, it still may be several years off before they can do it nearly as skillfully as humans. So your human customer service rep’s role may be more about calming an irate customer than solving their problem. The problem solving part will be handled in real time by the rep’s automated partner. 

Adapting to this change in roles for both humans and their automated assistants will be a major challenge for companies over the next couple of years. Interesting times ahead!


0 comments
96 views

Permalink