What I saw at Knowledge18

Scenes from the future of customer service

Artificial intelligence in customer service
  • Service management changes the division of labor between machines and people, but isn’t a replacement for either
  • Delivering great digital services requires connecting data silos
  • Process automation software helps Tennessee dig out from crisis

I just got back from Knowledge18, an annual tech conference sponsored by ServiceNow, the publisher of Workflow™. Some 18,000 IT pros, analysts and journalists trooped to Las Vegas to check out new ServiceNow products and mingle with their peers.

ServiceNow sells cloud‑based software that automates repetitive tasks inside organizations. The industry term for this technology is service management. It covers pretty much all the necessary but mundane processes that make you want to head‑butt your keyboard—things like changing a computer password, onboarding new employees, prioritizing and routing customer issues, and mitigating security breaches.

Essentially, service management changes the division of labor between machines and people: Intelligent machines take on tasks that can be broken down into a series of repeatable steps, freeing up people to use uniquely human traits like creativity and empathy to solve non‑routine problems.

Customer service is a case in point. How do you deliver great services to your customers in a connected world where people expect instant, 24/7 solutions to their problems? The key is delivering services that are “effortless, connected and proactive,” according to Abhijit Mitra, general manager of ServiceNow’s customer service management business unit.

Take common customer experiences like returning a retail purchase or refilling a prescription. In his Knowledge18 keynote, Mitra said you can “automate these requests as workflows, and publish them as self‑service options on your website. So your customers don’t need to call you anymore. They can simply go to your website and with a click of a button, the services are instantaneously delivered to them.”

Behind the scenes, organizations need to connect silos so that teams can work across functions. “Many times, the people involved in the process of delivering services to customers are spread across multiple departments, like engineering, operations and finance,” Mitra said. “So you need to get everyone together on a common platform, assign them work, and track that work to closure. And now, customer service becomes a team sport.”

Proactive service is the final leg of the stool. According to Mitra, it’s about “monitoring data, analyzing trends, even predicting issues, and fixing them before customers are affected.” And not just historical data, the kind that traditional CRM systems capture. “It could be data from the Internet of Things that’s analyzed in real time to avoid outages and prevent business disruptions.”

One of the keynote presentations at Knowledge18 featured a demo built around a fictional autonomous lawnmower called LawnBot. Due to a software bug, LawnBot has developed the unfortunate habit of chasing cats up trees. “That’s a great YouTube video but a terrible customer experience,” said presenter Michael Fortson.

Fortson is the product manager for Virtual Agent, a new ServiceNow product that uses machine learning to automatically categorize, route, and prioritize customer issues. It learns from patterns in your historical data, becoming increasingly accurate in its predictive recommendations.

Instead of calling customer service to help tame the rogue machine, you can simply visit the LawnBot website and chat with an intelligent bot that knows who you are, identifies your problem, and solves it by automatically rebooting your LawnBot. (Virtual Agent was announced at Knowledge18 and will be built into future releases of the ServiceNow Platform.)

Prefer real examples? During my time at Knowledge18 I stopped by a presentation on customer service in government. It was led by Landon Cook, the director of customer service operations at the state of Tennessee’s Department of Human Services (DHS). Cook’s department delivers services to Tennessee citizens with serious, real‑life issues. Some of them are trying to get back into the workforce. Others have disabilities or are living in poverty.

Customer service loomed large at the Knowledge18 conference in Las Vegas

Last year, DHS experienced what Cook described as “the worst customer service delivery situation I’ve ever seen.” The department had instituted an early retirement/buyout program so that it could then hire younger employees. Many of the people who took the buyouts were customer service reps working in call centers. Soon after they left, the state announced a hiring freeze, which meant that DHS couldn’t replace them.

At which point, DHS experienced an influx of new customers. “It was the perfect storm,” Cook recalled. “Wait times for calls went from mediocre to abominable.” How bad was it? The peak phone wait time reached two hours. Email response times hit an average of 1.5 days.

“We in customer service became the bad guys,” Cook said. Cook worked with ServiceNow to implement a customer service management system that included automated direct email with powerful reporting capabilities that dramatically reduced inquiry duplication and increased collaboration across DHS program areas.

DHS spent two years planning the new system, and just six weeks rolling it out. The results were dramatic. The average time to resolve an escalated customer inquiry fell from 80‑plus hours in September 2017 to less than 35 hours in November. “We went from being the Fred Flintstone of customer service to the James Bond,” Cook said.

The new technology allows DHS customer service professionals to concentrate on solving customer problems instead of entering data in spreadsheets. “They didn’t join DHS to do data entry,” Cook said. “They joined to serve customers.” Thanks to service management, they can now do just that.

These stories point to an emerging paradigm for how humans and machines relate to one another at work. Despite widespread anxiety that robots will eliminate human jobs, what I glimpsed at Knowledge18 is a world where humans and machines are more effective as partners than as competitors.

For that vision to become reality, organizations will need to get very good at supporting networks of human and nonhuman workers. In an appearance on the main stage at Knowledge18, Accenture CIO Andrew Wilson noted the irony: “In an age of ultra‑fast technology,” he said, “humans are more important than ever.”