Reskilling the Contact Center for the AI Age

Artificial Intelligence could be the technology that breaks that tendency. Unlike past forms of automation, AI will become ever more capable of tasks requiring that most human of qualities, insight.

As we’ve explored previously, AI will change the contact center in profound ways. In this piece, we review the short and medium term impacts that AI is having on the contact center, and then we address what that means for the roles we need to staff those contact centers.

Why Might it be Different This Time?

Some people fear that AI poses a different risk for jobs than previous technological innovations.

We can see why by looking at the Luddites.

They fought against steam-powered looms, fearing those machines would take the jobs of hand weavers. However, powered looms not only automated manual labor but made weaving more efficient. In America, powered looms increased a single weaver’s productivity by 50 times. That meant the price of cloth fell and people could buy more, which in turn enabled weavers to sell more.

AI seems different. AI automates mental labor, rather than manual labor, and it’s not immediately clear how an oncologist, for example, benefits from an AI that makes better diagnoses and recommendations than they could.

This strikes at our sense of what it means to be human; we’re meant to be the intelligent ones, after all.

So, the first challenge of introducing AI to the contact center is to help our colleagues to understand what it means not just for their immediate careers but also for their place in the world.

AI’s Role in the Contact Center Today

Back in 2011, Gartner predicted that by 2020, 85% of customer interaction would be without the aid of a human on the business side. With just 14 months to go, how’s that looking?

Today, AI’s role in the contact center falls mainly into three categories:

• Analysis of past activity, such as software that can analyze every single call, email, tweet, and SMS that your contact center receives.
• In-the-moment assistance, whether that’s a bot interacting directly with customers or an assistant that aids contact center staff as they interact with customers.
• Prediction, from helping contact center managers to schedule staff appropriately to identifying when individual customers are likely to churn, complain, or upgrade.

Each of these augments, rather than replaces, human agents.

Communication Analysis

Analyzing recordings of customer calls, for example, has long been a task carried out by training or quality assurance managers. But it’s time-consuming and humans struggle to draw useful conclusions from qualitative data.

By using AI techniques, such as natural language understanding and machine learning, software can analyze entire customer contact histories in the time it would take a human to listen to one or two recorded calls. What’s more, the computer can dispassionately identify patterns and then draw data-driven conclusions from them.

In-the-Moment Assistance

When it comes to in-the-moment assistance, AI and self-service combined are helping contact center agents to move up the value chain. As we build products and services that increasingly meet customers’ desire to self-serve, contact center agents are no longer needed to answer basic factual questions. “What’s my balance?” can be answered with an app or a chatbot.

Similarly, AI-powered virtual agents can listen in on conversations and provide timely assistance to the human agent. For example, if an auto insurance customer calls to make a claim, an AI-agent could listen for details and pull up pertinent information, such as the traffic and weather conditions at the scene of the accident or even live video from CCTV.

Such details could help in establishing the circumstances of the incident or in judging the need for emergency or recovery services. Using cloud communication APIs, such as Nexmo’s Voice API, developers can piece together real-time data, AI-services, and the customer’s call to enable the human agent to focus on delivering the right level of care and support.

Prediction

Reliable forecasting––of demand, customer behavior, agent performance––are essential to the efficient operation of a contact center.

AI-powered tools can use past data to predict likely future outcomes. To return to our example of the auto insurer, an AI-powered staff rostering tool might recognize that claims increase in the 48 hours following heavy rain and so use weather forecasts to recommend that management increase the number of agents working during that time.

AI in the Near to Medium Term

Over the next decade, several strands of technology will converge to continue the trend we see today where the contact center is becoming less of a physical place and, instead, more a function spread across locations and technologies.

Cloud communication APIs are already eroding the barriers of traditional telecoms and are enabling true omnichannel communication. Chatbots are finding their way into frontline customer contact. AI-powered tools are improving the efficiency of human agents.

None of that means that we’ll be able to remove customer service staffing from budgets. But it does mean that AI is following the pattern of previous industrial automation: it is freeing humans from low-skill work and helping to make higher skilled work easier and more consistent.

In five to ten years, we’ll see increasingly capable bots handling phone calls, social media, and chat inquiries. Crucially, they’ll seamlessly escalate queries, when necessary, to human agents who will be aided by other AI-powered tools.

It’s foolhardy to make predictions about technology; but writing in 2018, it seems unlikely that the contact center of 2028 will have virtual agents that are as capable as humans. But might they be good enough in most cases?

New Roles, Same Job

AI won’t fundamentally change the job of the contact center as a whole but it will change who works there.

Opportunities for lower-skill, frontline agents are likely to diminish as AI-powered virtual agents become more capable and more consistent. But a whole new tranche of job roles will arise in three categories:

• Keeping the AI running
• Performing essentially human roles that AIs cannot (yet)
• Finding the best ways to make use of the AI’s capabilities.

The first category –– keeping the AI running –– is essentially a function of the IT team. Developers will produce AI capabilities and integrate them with customer facing channels through cloud communication APIs.

The second two categories, though, are as much an embedded part of the contact center of the near future as a phone agent is today.

Human Roles

As customers come to prefer self-service and, perhaps later, using virtual assistants to interact with your organization, what role remains for humans?

According to the American Express Customer Service Barometer, 68 percent of Americans say that “pleasantness” drives a great customer service experience. While it’s a much smaller percentage, 18 percent of Americans say a personal connection is required to improve customer service. And, perhaps most important of all, a person’s desire to speak with a human increases with the complexity of the issue.

AIs cannot yet make a convincing personal connection; and perhaps we’d be wrong to think they should. After all, customer service is one of the battlegrounds on which companies fight for customers.

The contact center of the near future needs to retain its most empathic, skilled agents to deal with those more complex issues. Where a customer feels something of value is at stake, contact centers will need a swift escalation path from bot to person.

People Supporting the AI

However, it’s in making the most of the AI-powered tools that we’ll see most new jobs appear in the contact center. Roles that might have existed elsewhere in a large company will become crucial in the contact center itself if we’re to ensure that AIs are helping us to provide excellent customer service rather than only making savings.

In particular, it’s likely that cutting edge contact centers will start to recruit for roles such as:

• Anthropologists, linguists, and psychologists: If AIs are to replace some forms of human contact, then we must make sure that we have a thorough understanding of what customers need and expect from their communication.
• Community managers: This is a role that already appears in customer service organizations, primarily looking after social media interaction, but it will become more important to have someone on staff who can understand the dynamics of remote communities. What does this have to do with AI? Non-AI interactions will hold so much more weight as they’re reserved for more important, more complex queries. Understanding and shaping the way that our customers interact with our agents, our AIs, and each other will be essential to maintaining market-leading customer service.
• Data scientists: Analyzing and making decisions based on the data generated and prepared by AIs will remain a human function. Data scientists have the expertise to cut straight to the right data and recommendations produced by AI-powered tools.

The contact center will feature an undoubtedly different roster of roles and capabilities than we see today. However, it will remain an ultimately human experience: humans will seek out humans when they need them. What changes is that AI will make contact center staff more efficient. Just as every other automation technology has––on the whole throughout time––AI will empower people rather than diminish them.