6-minute read

Quick summary: How advanced uses of call center AI are helping organizations improve customer loyalty, reduce per-contact costs, and decrease call center attrition rates

The evolution of AI customer service bots into intelligent virtual assistants (VAs) has changed the customer service game to benefit both consumers and businesses. Driven by advances in technology and escalating expectations of their customer service centers—both on the consumer side and among call center reps—organizations are leveraging automation and machine learning to provide their customers with efficient, accurate solutions while also reducing per-contact costs.

Chatbots have been around for decades (the first was ELIZA in 1966 at MIT), with varying degrees of success. Early efforts using conditional logic (“if this, then that”) and keyword matching (generating scripted responses based on keywords in user inquiries) performed reasonably well in answering simple questions; however, they lacked the ability to predict customer intent and to engage in meaningful, dynamic dialogue. Fortunately, advances in artificial intelligence techniques like machine learning and computer vision are transforming our experiences with virtual assistants by effectively mimicking human interaction while addressing complex issues with some of our best traits: empathy and kindness.

Learn more about conversational AI for customer service

The semantic difference

Today, artificial intelligence for customer support readily leverages the latest in semantic understanding technology to better understand the user’s input, locate the source of needed information, and deliver an accurate, timely response. The way a baby’s mind learns meaning in languages, beyond word selection and connotation, is inspiring the way VAs are built. By developing self-learning, customer intent recognition, and EQ-demonstrating solutions, it is possible to create virtual assistants that think and talk like the most skilled, qualified human agent. The result: customer experiences with rapid issue resolution across multiple channels (mobile app, web, voice) and subsequently increased self-service and CSAT KPIs.

The evolution of artificial intelligence for customer support

While earlier chatbots used basic algorithms to deduce meaning, advanced AI allows VAs to predict and more accurately identify the intents behind user requests, and to integrate with enterprise systems of record to formulate responses. This allows VAs to better personalize and respond to customers during interactions, and it allows the customer’s data (which may be stored on CRM, ERP, HCM, and ITSM systems) to be leveraged more effectively.

Furthermore, artificial intelligence for customer support now enables VAs to learn from each interaction to inform future conversations. If transcribed sentiment data indicates that a message causes confusion or frustration among users, the virtual assistant can adjust the message to improve accuracy or to deliver information in a more accommodating tone. Machine learning also enables VAs to connect user inquiries with personalized upselling and cross-selling opportunities.

Example: When a passenger asks an airline VA “When will my flight get in?” the virtual assistant understands the entry as a request for an ETA, requests and locates the user’s record (perhaps automatically, using computer vision and face recognition), along with their flight information, and responds in real time, accurately, and in a secure, compliant way. Or, if a user complains about a certain aspect of a flight, the VA can, after analyzing possible causes of frustration, offer some form of recompense based on the customer’s situation, loyalty, etc.

Sentiment analysis

Advanced call center AI enables virtual assistants to pick up on cues to a user’s frame of mind and deliver the appropriate, high-EQ responses to customers who are worried, angry, frustrated, etc. Two different customers, for example, may be reporting the same issue — but one seems to be indifferent while the other is clearly angry. The VA may use a standard conversation design to respond to the first, while it may choose to seamlessly transfer the second to a human agent, who can address the concern personally and smooth things over with the customer. Most importantly, all the data from the customer’s interaction with the virtual assistant will be preserved for the human agent, ensuring there will be no surprise “start-overs” for the angry customer.

Example: A user who says “I get an error message when I try to check out” may be led down a standard self-service dialogue to troubleshoot issues; however, another who says “I’ve tried to check out three times and keep getting an error message that doesn’t make sense” may be connected directly to a human agent, not only because of what was said, but how it was said.

Why virtual assistants?

Virtual assistants can help organizations realize millions of dollars in savings, double-digit growth in Net Promoter Score (NPS), predictive real-time intent prediction, and valuable analytics insights, just to name a few benefits. Here are three specific ways VAs can substantially improve the bottom line while modernizing the whole experience:

1. Customer self-service

Not only are today’s consumers comfortable conversing with virtual entities—thanks to Siri, Alexa, and their cohorts—but many prefer virtual interactions to human conversations in customer service situations. In a recent study, 59 percent of consumers surveyed said they prefer service that doesn’t involve speaking with a call center rep, and 33 percent said they would like to see all customer service go through virtual assistants.

2. Call center efficiency: handling growing volume with lower costs

Businesses facing considerable competitive pressures are challenged to balance innovation, customer satisfaction, and costs, especially in an area where all three coincide: their call centers. According to Gartner, real-time customer service channels involving humans—such as phone, live chat, and email—cost on average $8.01 per contact, while virtual interactions for simple queries can run as low as $0.10 per contact and have proven to deliver savings at scale. By taking on requests and inquiries that don’t require human interaction, call center AI can directly reduce costs and improve quality of service.

3. Call center employee retention

Employee turnover rates in call centers are unusually high compared to other departments—between 30 and 45 percent on average—particularly due to monotonous, simple, and repetitive requests. Loss of good agents negatively impacts morale, and the cost of training their replacements averages about $7,500 per employee. By automating simple tasks that are better accomplished using a virtual worker—like helping customers change passwords, check order status, and reschedule deliveries—call center AI leaves human employees free to help customers with more complex, interesting challenges. Human agents can focus on areas where they add the greatest value, such as focusing on fringe cases, putting out fires, making connections that a robot can’t, and offering the priceless human touch.

Use cases for call center AI

Call center AI presents a broad, but solvable, set of interactions that can easily be modeled. Below are some prevalent examples of areas where organizations are meeting the demands of growing volume in a cost-effective way using artificial intelligence for customer support:

  • Placing an online order
  • Updating a phone plan
  • Booking flights
  • Applying for a new credit card or mortgage
  • Filing an insurance claim
  • Narrowing down product choices
  • Seamlessly transferring to a human agent
  • Intent-driven engagement to deepen relationships
  • Reducing internal costs for employee onboarding, IT support, benefits administration, etc.

The future of artificial intelligence for customer support has arrived

Advances in technology have created new possibilities for engaging users through intelligent call center AI, while growing consumer expectations of instant, personalized, virtual interactions have opened new opportunities. Organizations that leverage these powerful solutions position themselves to improve customer loyalty, reduce per-contact costs, decrease call center attrition rates, and streamline operations to scale efficiently. As virtual assistants continue to become more intelligent and capable, new and more complex use cases will continue to emerge, which will revolutionize the way service is delivered globally across an organization, impacting customers, employees, and everyone in between.

Like what you see?

Paul Lee

Amit Unadkat is a Senior Manager of Digital Transformation with extensive experience in robotic process automation, virtual assistants, business process optimization, and technical product management. In 2021 he received Built In’s Tech Innovator Award for his work in automation and was recognized as a Rising Star by Consulting Magazine.