Tim Joyce, CIO at WDS, A Xerox Company

“The ‘intelligence’ part of AI is actually supplied by customer-service agents.” — Tim Joyce, CIO at WDS, A Xerox Company

By Tim Joyce

We’ve been talking for decades now about artificial intelligence (AI) and how it might help make our lives easier. How it might free us from the drudgery of repetitive work and automate tasks. Give us better insights and help us do our jobs better. And maybe, just maybe, one day watch what we do and learn from us.

The really exciting thing is that when it comes to customer care, we’re actually there. Right now. And what’s even more exciting is that it doesn’t spell the end for humans – far from it.

Because the “intelligence” part of AI is actually supplied by customer-service agents. They’ll continue to play a vital role, as they keep the machines on track through ‘guided learning’ and make sure they keep pace with the ever-changing nature of contact centers.

And here’s the really clever bit: all that intelligence is then used to create a pool of knowledge that helps other agents deliver faster, better service. It’s also used to automate self-service systems.

The Four Stages of Customer Care
A more sustainable, more value-generating (and more exciting) way to do customer care going forward. Read this eBook from Xerox.

But let’s take a step back here, and see how it all works in practice:

  • A customer-service agent takes a call and helps a customer with a problem. They work through steps in sequence and eventually determine what the root cause is and how to resolve it.
  • All through the process, the system is learning: Watching each step, following the sequence, seeing what works and what doesn’t, and determining the shortest path to resolution from the starting point.
  • The next time a customer-service agent handles a similar call, the system recognizes the problem and can cut to the chase, discarding false trails and unnecessary steps.
  • All of this intelligence is also fed into self-help systems, so that in future, the issue can be handled automatically, without any human intervention.

The result is faster, better service as the system learns from interactions, making each one better than the previous, streamlining processes, and constantly reducing the number of steps to get to a resolution.

It almost seems paradoxical: lower costs and better service. But it’s true. And it creates a virtuous circle:

The service agent is augmented by machine learning, resulting in faster resolution, shorter call time, lower costs and higher customer satisfaction.

Machine learning is enriched by each interaction, creating a vast repository of knowledge that can be tapped into by agents and used for self-service customer care.

Maybe you’ve spotted the obvious cloud on the horizon, though.

The Robots Are Coming

If machines are so good at learning, are we facing the nightmare scenario of them putting humans out of a job? Gartner reckons that by 2020, 85 percent of transactions will be handled without human interaction. So where does that leave humans? Does mass automation mean mass unemployment?

In a word, no.#WorkingBetter

Customer-service agents will simply focus on different things, as their roles change. They’ll become intelligence teachers, as they take transactions and turn them into learning for the system. They’ll also be community managers, monitoring social-media forums, proactively fixing problems and defusing potentially damaging situations.

And preventing those all-important customer defections. When a 2 percent increase in customer retention is equal to a 10 percent cut in costs, that’s no mean feat.

It’s important to remember that more complex issues will always be handled by people, as the routine, repetitive tasks are handled automatically. Some stuff a machine can never sort out, and reassurance is always better coming from a real person. And sometimes, that’s all your customer needs to remain your customer.

With this new, more interesting role, customer‑service agents will feel more challenged, more rewarded and more valuable. Churn rates will be lower and real ‘tribal knowledge’ will be built up both among call-center staff and the machine-learning system.

Back to the Future

Customer care is at a crossroads, where it can carry on what it’s been doing for as long as anybody can remember – or it can change. And that means blending seats & software, which is now the new normal.

Forty-two percent of queries currently can’t be solved by customer-service agents because of inadequate systems, according to Forrester. So it’s crucial to make sure that people and technology are connected, integrated and seamless. And that you get the best from both, to give the best to your customer.

These are exciting times indeed for customer care, with technology pushing the boundaries and making the impossible possible, as we head towards an unpredictable future.

Subscribe to this blog and receive email updates when we publish a new article.