When it comes to automated customer service, there’s a lot of news. We’re doing a fact check to let you know what’s hype and what’s possible today.
The evolution of AI is pretty remarkable. Just look at self-driving cars. While it is astonishing how far the technology has come and the pilot programs that are underway, the days when we’re all watching our favorite shows while our cars whisk us to work are still a few years off. If you read the headlines, though, you might think we’re either going to have fleets of autonomous cars on our highways tomorrow or self-driving cars will end humanity as we know it.
As with many things these days, there’s a lot of polarizing fake news circulating around AI, including with the application of automated customer service. Some reports and companies have over-hyped the capabilities of AI promising solutions that are not capable of today. Others have issued scare tactics akin to the Hollywood portrayal of artificial intelligence, shunning the technology in the name of job loss or the ability for AI to provide a meaningful impact.
Let’s fact check some of the most prevalent things we hear:
HEADLINE: AI can automate ALL of your customer support queries
FACT CHECK: AI can automate your high-volume, repeatable queries (~35%)
AI for customer service can effectively resolve repeatable, simple queries like order status, account/subscription updates or refund requests. By integrating with core systems, like CRM and OMS, an automated agent can provide replies to customer queries in a few seconds. You should not try to automate complex or emotionally-charged topics. The queries that require some level of objectivity or empathy need human oversight.
HEADLINE: Your customers hate bots.
FACT CHECK: When deployed correctly, customers appreciate the convenience and instantaneous results that bots can provide.
This one is tricky. Consumers are excited about getting personalized, instantaneous support that is made available with AI. It’s fair to say, though, that most people feel shortchanged by the interactions they’ve had with customer service bots to date, noting they are not as good as humans at accomplishing things or have not been given the authority to make decisions.
Automated agents need to have the authority to resolve tickets; empowered to approve or reject a return request, upgrade someone on a flight or refund someone under a certain price point. Make sure you always give customers an “out” to speak with a human in every circumstance.
HEADLINE: AI can predict customer needs.
FACT CHECK: With the right integrations, AI can help deliver proactive customer service.
While full-blown predictive service capabilities are still being developed, proactive service is a reality. if you integrate your AI platform with your core systems of record, you can proactively reach out to customers with important information. For instance, if a sweater comes back in stock that was sitting in a person’s cart, proactively send them a message to know it’s now available. If a snowstorm will cause a package to be delivered outside of the promised window, you can automate a message to the customer to let them know.
HEADLINE: AI will wipe out customer support jobs.
FACT CHECK: AI needs to collaborate with human agents to provide frustration-free service
There’s a clear division of labor between human and AI in customer support. AI needs to be deployed to solve repeatable queries, while uplifting human agents to resolve the rest. AI can collect information from the customer or other business systems to package relevant information for a human agent to review and make a decision. AI leads to happier employees, as their jobs become more fulfilling (more high-impact work, less mundane tasks) decreasing agent turnover.
In summary, AI is advancing at a pretty incredible rate….. and the industry is still maturing. AI needs to be deployed within contact centers correctly to solve the right issues (highly repeatable), leverage natural language understanding to empower customers to engage naturally and connect to the right internal systems to have the authority to resolve.