The worker who is closest to the customer is the one who knows the most. Yet this truth is discovered and forgotten with alarming regularity and so frequently underappreciated.
So underappreciated, in fact, that we are seeing an ever greater push towards automation of the customer service experience – a trend that is largely serving the interests of hands-off leaders, rather than those of their actual customers.
Outsourcing your customer service interface is often a poor decision.
Just think of the variable experiences that people have with delivery drivers.
You can have great customer service over the phone from the primary vendor. You can buy a quality product. You can be treated well.
But if it’s delivered by people who don’t work for the main vendor, who are not part of the same culture and don’t subscribe to its values, then the risk is that your initial swell of goodwill will simply break down.
Or even go into reverse.
AI tidal wave
Issues around automated customer service have surged back into the spotlight, thanks to NatWest unveiling its prototype virtual bank teller, Cora.
While clearly a sophisticated piece of software, Cora has already proven in test runs to have trouble with processing certain accents and customer queries.
Yet Cora heralds what promises to be an artificial intelligence tidal wave that is set to roll across the customer experience in the coming years. Technology market research giant Gartner predicts that, by 2020, a quarter of the world’s customer-service departments will deploy ‘virtual assistants’.
However, at the point where customer service is delivered, we want something that is efficient, gives us solid information and doesn’t keep redirecting us time and again.
There can be little doubt that automated phone lines – “Press four for this and five for that” – are one of the least appealing aspects of modern life. We like to hear a person on the other end of the phone. This is not merely anecdotal: there is hard evidence that we overwhelmingly appreciate a human touch.
When there’s a need for empathy and sympathy, how can AI reproduce that?
Emotional response
Last year, I attended a conference on AI technologies in Iceland. I was eager to ask some of the experts in attendance when we are likely to have a robot that can detect an emotional response and modify its own response accordingly. A professor I spoke to said it’s still some way off.
Organisations that are thinking about implementing these tools must be very careful and discerning about which parts of the customer service experience are typically fast, efficient and rote – and are therefore safe to automate – and which ones depend upon a more sustained, empathic, human interaction.
A man by the name of Roger Jones – who came up through the trade union movement and eventually joined the faculty at Cranfield School of Management – developed a model called the Four Orders of Administration.
Jones proposed that, in ‘first-order administration’, anything to do with the customer is vital – and the more removed you get from the customer, the less use you are. By the time you get to the fourth order, he added, you’re contributing almost nothing to the customer experience.
The sting? That fourth order routinely encompasses the C-team.
So, the essential message is: move your staff away from your customers at your peril.