How Well Can Artificial Intelligence Run the Help Desk of the Future?

By Tim Joyce, CIO at WDS, A Xerox Company

(From the editor: This article was originally published on The Huffington Post.)

It’s a very hot day. Your calendar is very tight, and the back of your mind is filled with anxiety about a sick child. You’re squeezing in lunch with a friend who just lost her job. You hold your hand over the biometric payment reader only to learn that credit has been denied. Your jobless friend has to pick up the tab.

42% of customer care agents are unable to efficiently resolve customer issues
Source: Forrester The next-generation contact center 2012)

On the way out, you contact your bank. You opt to speak to an agent rather than using a text-based alternative. The sun is blazing, and you feel your blood pressure rising.

The agent who answers apologizes for the mishap before you say a word. She already knows your name, and suggests you step into the shade while she inquires about the lunch bill. Within seconds she’s discovered the problem and offers a solution. Her confidence and efficiency is soothing – she’s  all business but also understanding. Before hanging up, she asks if you’d like to have her credit your friend’s account for lunch – a nice touch. This is the only question she asked during the entire 60 second conversation.

Compassion from a Machine?

The agent is not human. The compassion and efficiency offered are based on data, not emotion. Its “personality” has been programmed to reflect the type of bank the agent represents – efficient, understanding, confident.

As software advances, the sophistication and perception of virtual customer care agents will radically change. Computers will fully emulate human beings. They will recognize emotion and mood. They will be able to dialogue with people in an intelligent manner. They even will be able to influence us with persuasive arguments.

Through these virtual beings, brands as we know them today also will become more “human.” Today, a brand’s voice relies on messaging, employee training and consistency. Tomorrow, these things will be embodied in a highly developed, adaptable software program. The various brand “beings” will be diverse and interesting because they will help brands differentiate themselves in a new economy that is based on service rather than products.

Service Will Define Your Products’ Success

If you think about it, business success stories during the past 100 years primarily were about products. We’re moving toward a world where success and value propositions will be based on service. We will need products, but our purchases are more about the services wrapped around them. Consumers will be permanently connected to sophisticated helpdesks that watch and anticipate customer needs.

This economic model will cause brands to develop virtual beings that serve customers creatively and in the best way possible. Virtual agents will be quite proactive, relying on data analytics to understand you. They will know the best time to talk to you if a problem crops up. And because they have so much information at their “fingertips,” they can talk you steadily through a difficult problem – in a manner that best suits your personality and knowledge-level.

A Vastly Different Role for Humans

Will human agents still be in the mix? Yes, but their role will be vastly different from today. As customer experience managers, they will govern the brand’s customer care architecture, drawing upon trends highlighted by their virtual counterparts to maintain the feedback loop and action the necessary changes to better serve customers and better reflect the values of the brand. It will be their job to ensure their virtual counterparts behave and evolve appropriately, so that every customer experience can be better than the last.

On that hot day when you were exasperated at lunch, the virtual agent was detecting emotion through the pitch and pace of your words, and had the ability to truly understand and empathize with your situation. It also overlooked your rudeness and maybe even offered a small joke to send you on your way after the problem was solved.

Tim Joyce, CIO at WDS, A Xerox CompanyXerox’s Tim Joyce spends his days thinking about conversations; both the individual conversations brands have when caring for their customers, and the big conversations” companies have as they consider markets and customer demographics. He wants companies to have better quality and more frequent communications with the people who choose to spend money with them. Officially, Tim serves as chief innovation officer in Xerox Customer Care where he deploys technology that helps millions of people have a better customer experience every day.

 

Subscribe to Simplify Work and receive email updates when we publish a new article.

We asked Xerox people: How do you feel about talking to computers (like Siri)?

 

Results: How do you feel about talking to computers (like Siri)?

 

Imagine you call customer care and a computer answers.  How will tomorrow’s computer need to behave to be as good or better than talking to a human?

Many of those who responded to our survey indicated that they “don’t believe that any machine now or in the future could be better in communication than humans.”  Some even expressed dismay at the idea.  “I realize it is old fashioned, but do we really need to dehumanize ourselves by talking with machines?”

By 2020, customers will manage 85% of their relationship with brands without interacting with a human.
Source: Gartner Customer 360 Summit 2011

But others pointed to potential benefits that only a computer could provide.  “It hopefully will ‘remember’ what you have said to it previously, so you don’t have to repeat yourself.  You probably will not need to be placed on hold” or ”transferred three or four times to get your issue resolved.”  Computers could more fully “listen to what is being said” because they “will lack self-interest” and won’t be “distracted.”  Software allows for a vast array of additional capabilities.  Computers may someday be “multi-lingual” and able to access databases of usage history to “anticipate every possible outcome.”

Still, there’s a long way to go before this can become a reality.  “The computer would have to be able to understand a problem from a description rather than playing yes/no questions for 15 minutes.” And then “The computer will need to be able to make decisions” and “negotiate with me.”

“The biggest challenge will be emotion.”   We’d need to program computers to “sort through the intangibles of voice – dialect, idioms, humor.”  To equal person-to-person dialogue, computers must “read between the lines” and “laugh at a joke.” “It needs to make me feel welcome and listened to. Appreciated. Ever feel appreciated as a customer by a computer? Not yet.”

But in the end, there is only one true test:  “I should not know I am talking to a computer.”

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

Related Posts

2 Comments

  1. Clayton MacLean March 17, 2015 -

    So computer automated helpdesks will score high on efficiency but not as high as one would expect when it comes to overall customer satisfaction with the service. Perhaps for the more complex issues where human understanding and compassion are needed then having a human being on the end of the line can never be equalled by an A. I.

    • Gregory Pings March 17, 2015 -

      Thank you for sharing your thoughts, Clayton. I passed them along to Tim Joyce; his reply follows:

      Automated helpdesks will improve efficiency and, I believe, customer satisfaction. Automation will see customers interfacing with pools of data, capturing everything they say to eradicate the poorer aspects of support. Imagine; customers will no longer have to endure frustrating transfers to different departments, or be asked to repeat their question as the data required will be surfaced to the customer interface as and when required.

      Human agents will always have a place in the customer care sphere, reimagined as specialized customer advocates, and they will be instrumental in resolving the newest and most challenging issues. In fact, these CSRs will be pivotal to the performance of automated customer care, as the A.I. will learn from them to duplicate their efforts on a larger scale.

Comments are closed.