WSJ's The Experts

How AI Can Make Your Call-Center Experience Less Painful

How AI Can Make Your Call-Center Experience Less Painful

AI-driven technologies power the engine for "predictive routing" to the right customer-service agent.

  • 읽기 소요시간


How AI Can Make Your Call-Center Experience Less Painful

This article originally appeared on WSJ's The Experts.

“Your call is important to us...” has been a running joke for years. While service centers have adopted technologies like callback options and voice recognition, once the customer reaches an agent, odds remain high that the problem won’t be solved in one go. That’s because the traditional first-in, first-out customer queue is fairly dumb about customer/agent matching. As a result, problems unsolved during the initial call lead to more complaints and wasted time as employees try to calm upset customers.

Artificial intelligence is here to help—really. Early efforts are enabling companies to improve the overall experience while reducing costs (in staff time, field technician visits and defecting customers) in the bargain.

AI-driven technologies power the engine for “predictive routing” to the right agent. For instance, AI might pair customer Maria directly with agent Alexander. How? AI blends what it has learned about Maria’s individual characteristics and history with a rigorous assessment of Alexander’s and other agents’ profiles. Maria might have made contact twice in the past week about a bill, or she might have been scouring real-estate and mortgage-comparison sites—information that AI can scrape from the company’s own databases and external websites.

On the agent side, AI compiles and measures the eventual outcomes of interactions in similar situations. By analyzing the data, it can understand what Alexander is good at, the problems he can solve and the problems he hasn’t excelled at solving. That’s a huge leap from how many companies traditionally have assessed their agents. They’ve kept spreadsheets listing each agent’s education, certification and presumed skill level based on training and tenure—none of which accurately predicts how well someone can handle specific types of interactions.

Based on all of this data, updated in real time, the routing engine matches Maria with a preferred agent, within the constraints of agent availability at that moment.

To continually improve the algorithms that predict outcomes of customer/agent interactions, machine-learning techniques flag certain patterns of customer behavior. For instance, multiple contacts by a customer may signal a propensity to cancel a contract; a call after a recent bill could mean the customer will dispute a charge; or the first inquiry after a bank sends a credit card to a customer likely will be to activate the card.

There’s more to this process, though, than just installing AI technology. Knowing that people value their time, companies have to nudge some customers to wait a bit longer for the right match. Many customers are willing to hold or get a scheduled callback if the company signals how that will better resolve their issue. Success with AI also requires changes to agent selection, training and scheduling. The human element will remain essential to solving customers’ problems.

Over time, AI-powered predictive routing should make customers’ lives easier as it spreads to more types of interactions, more accurately predicts outcomes and speeds up matchmaking. When agents spend more time on the issues they are most suited for, they’re happier and sharper, which spills over into every customer conversation. Instead of feeling dread, you might even relish making that next—more successful—customer-service call.

James Allen is co-leader of the global strategy practice at Bain & Co. and co-author of The Founder’s Mentality.


베인에 궁금하신 점이 있으신가요?

베인은 주저 없이 변화를 마주할 줄 아는 용감한 리더들과 함께합니다. 그리고, 이들의 담대한 용기는 고객사의 성공으로 이어집니다.