Artificial intelligence has become common in many households and workplaces. Popular AI personal assistants like Google Assistant, Siri, Cortana, Alexa and Bixby have played a big role in rapidly reshaping consumer attitudes about how companies should respond to them. And customers bring these expectations with them when they interact with contact centers.
Most consumers enjoy the autonomy of helping themselves—and they want answers and service faster than ever. Today, live chat is the dominant contact channel for online customers. Forty-two percent (42%) of customers prefer live chat compared to just 23% for email and 16% for social media, according to J.D. Power. It’s not surprising since, as cloud CRM platform provider SuperOffice reports, chat response time is typically two minutes, compared to average response times of 10 hours for social media and 17 hours by email.
Complex New World
AI is already in many contact centers and is being used to help drive customer satisfaction. The technology can take many forms and its use is still evolving. While it brings benefits in some areas, it also brings new challenges.
Swedish bank SEB uses a sophisticated bot named AIDA to interface with customers. As reported in Harvard Business Review: “Able to handle natural-language conversations, Aida has access to vast stores of data and can answer many frequently asked questions, such as how to open an account or make cross-border payments. She can also ask callers follow-up questions to solve their problems, and she’s able to analyze a caller’s tone of voice (frustrated vs. appreciative, for instance) and use that information to provide better service later.” Despite these extensive skills, even Aida gets stumped, handing off roughly 30% of calls to human contact center reps for resolution.
What’s the effect of all the easy-to-solve issues going to bots? For human agents, it means their workload just got more difficult. Now they spend a higher percentage of their time dealing with escalated issues, interacting with callers who may be more frustrated or dissatisfied, and attempting to resolve complicated, multi-part problems or needs.
Stress on Agents
These changes come on the heels of the COVID-19 pandemic, which already increased pressures for contact center workers. Many faced greater workloads as e-commerce traffic soared, and some also faced multiple obstacles and challenges adapting to working from home.
As the pandemic lingered, contact center workforces have ebbed and flowed as some verticals lost business and others gained. New workers came into jobs, often being trained remotely, then virtually “nested” and placed into production. These are capable, vetted workers who demonstrate an ability to adapt to these new processes, yet they don’t have the advantages of the in-person workplace. They cannot do traditional shadowing, break up stress with a laugh or joke with colleagues at their side, or flag down a supervisor to problem-solve at a moment’s notice. In relative isolation, they work to resolve ever more complex issues, perhaps thumbing through a written help guide or urgently texting a supervisor for help. The result is often greater uncertainty, nervousness or anxiety as workers struggle to maintain performance with less access—not only to information but the sense of camaraderie and teamwork that humans thrive on.
We’ve seen an increase in call center turnover, perhaps driven by these increased stressors. As a help, some employers are using co-worker chatrooms for stress relief. They allow phone reps to stop in daily as needed for 10-15 minutes at a time for a facilitated chat where small groups (typically five or six people) can talk and share to provide empathy and support.
Pressures on Supervisors
The call center manager or supervisor typically is hired or promoted for demonstrating strengths in very human skill sets like empathy, listening and creative problem-solving, as well as for their knowledge. Fundamentally rooted in interpersonal communication, it’s a subtle and vital role that no machine can ever fully replicate.
Yet, however well qualified they were as proven agents, supervisors may struggle learning to translate their abilities into the new role. Recently promoted supervisors need training, which can be harder to provide to them virtually. And even seasoned supervisors need training and support to learn to adapt the skills they honed in the in-person environment, now that they and their teams are remote.
Supervisors also face pressures trying to provide coaching and problem-solving when outnumbered by new staffers, who typically graduate from training 20 to 30 at a time.
How can AI help alleviate the pressures call center agents and supervisors experience, including some of the issues it has caused?
Exploring the AI-enabled Workforce
An Accenture study looked at 1,500 companies around the world to see which AI strategies worked best. The clear outcome was that, while AI failed when sought as a replacement for human workers, organizations that found ways for machines and humans to work together saw significant performance improvements. “Through such collaborative intelligence, humans and AI actively enhance each other’s complementary strengths: the leadership, teamwork, creativity and social skills of the former, and the speed, scalability, and quantitative capabilities of the latter,” the authors noted.
AI can be part of the solution. One effective use is AI that listens and suggests potential help to phone agents. The system picks up on keywords and filters through answers, placing relevant material nearby. It prompts them much the way Microsoft’s personal productivity assistant Cortana does as it monitors email messages and calendar items. The approach helps drive efficiency, putting potential solutions at workers’ fingertips and augmenting the role that the human coach can play in providing a sense of being supported.
While these tools need to be fed data to work with (a human role, involving capturing knowledge and incorporating the wisdom of exceptional workers) and given specific tasks and parameters to follow, they then proceed independently over time. As they do, they continuously learn what works and get better at delivering information the agent needs to resolve issues for customers.
This issue of scalability is also one that machines can address. A supervisor is not able to be in multiple places at once, but an AI system can readily handle multiple challenges simultaneously—after it has been trained and provided with the information and guidelines it needs by humans.
Let’s Talk About Chat
We’ve noted above the limitations of chatbots in problem-solving, but the combination of a skilled human agent using chat—commonly referred to as “live chat”—can be significant.
Our observations over the past year align with the growth predicted by most experts. One client began the pandemic using only 10% live chat, while today a full 40% of all interactions are by chat. They completely shut down the email function.
Live-chat workers need different skills than phone agents. It’s not like casual texting; responses can’t be full of shorthand LOLs and such. Chat agents need to demonstrate strong grammar, typing skills (typically 40 words per minute and up) and pass tests on social media etiquette. They must be careful communicators who proofread their own work quickly and efficiently. While some chat platforms integrate emojis, even these must be approached with professionalism that reflects the company’s brand and tone.
Many traditional agents are expected to cover chat, yet not every good phone agent is also skilled at chat. As we’ve seen, the integration of new tools brings both new stressors and new benefits, and the way it plays out can be a result of many factors. These range from personal preference to whether the workplace culture understands and commits to the need to support call center agents and their supervisors in embracing change.
Solving the Tough Cases
The ability to resolve issues competently and quickly is a core skill set of the contact center worker. That principle will not change whether the channel being used is a chatbot, an agent providing live chat, or an agent working by phone in a corporate or home setting. Figuring out the right mix of tools can be a question that evolves gradually as a contact center experiments and learns. Sometimes having all the tools at hand is the best approach, since a complex call may get to the point where it is best handled by phone.
Phone makes it easier to express empathy and allow understanding and connection to spark between people. When a person, not a bot, says, “I understand you’re having a problem with X, is that right? Please tell me about it,” it can de-escalate and persuade effectively. Since there is no typing messages back and forth, there is no delay between each exchange—a welcome factor for an exasperated caller.
Ultimately, AI and automated tools can help humans and can learn from them how to improve its own performance. That SEG banking customer service bot that currently transfers nearly one-third of the calls it cannot resolve? After it refers the call, it also listens in as the human agent handles the situation, all the while learning better ways to problem-solve. Perhaps some day, it will handle more scenarios in ways that include what it learns from observing people. Or perhaps not. It may be that people always turn to people for certain things.
Today, it is the escalated, stressful calls that go to humans, but these are also some of the more interesting tasks—issues that require research, creative thinking and the ability to imagine and envision solutions. It may be a billing discrepancy for a medical procedure that requires coding research and reaching out to a hospital and other providers, going beyond expected departments or to sources outside the typical channels. People excel at this multifaceted problem-solving and creative thinking. Though challenging, these call types can also be among the most satisfying issues to resolve, both for the mental variety and for the experience of working directly with another person and making a difference for them.
Humans and Machines: Different but Complementary
As our world and our inventions continue to change, they create new challenges but also new opportunities to collaborate and grow. As the Accenture AI study concluded, “What comes naturally to people (making a joke, for example) can be tricky for machines, and what’s straightforward for machines (analyzing gigabytes of data) remains virtually impossible for humans.” These are two distinct types of intelligence and capabilities.
Call centers need both.