A person using an AI-backed customer service messaging platform
PHOTO: Alicia Christin Gerald on Unsplash

Call centers can use artificial intelligence (AI) to help customers in many ways, reducing resolution times for users and the amount of work required of agents.

Brands commonly use conversational AI to scale conversations with customers. Unlike one-to-one phone calls or static websites, conversational AI delivers personalized experiences for thousands — even millions — of people where and when they want.

This ability to scale has become more valuable as people continue to hold most of their communications via messaging conversations.

By combining conversational AI for messaging, speech recognition and analytics for voice and integrations for a holistic view of all conversations, brands get a more comprehensive, unified system to connect across channels and reduce consumer frustrations.

“Messaging and conversational AI leverage channels that consumers use every day to communicate with friends and family, be it Apple Messages, WhatsApp or Facebook Messenger,” Joe Bradley, chief scientist at LivePerson, told CMSWire.

Bradley explained that with conversational AI tools, contact centers can measure how they perform against resolving customer intents, or the reasons why they reach out.

For example, LivePerson’s Meaningful Automated Conversation Score can provide information on where bots perform well (or not) at each turn of the conversation, allowing organizations to drill down and discover how to optimize customer conversations continually.

AI Can Augment the Human Touch

Poorvi Shrivastav, GM and vice president of products at HubSpot, pointed out that AI is used throughout the entire lifecycle of a call to reduce operational overhead and provide a personalized customer experience.

“Call center leaders can leverage AI to analyze customer sentiments, agent performance and key problem areas in order to create a stronger support experience,” she said.

Additionally, AI can replace repetitive tasks, allowing agents to focus on the root cause of most customer challenges instead of getting bogged down by administrative tasks.

“We’ve all experienced the frustration that comes with long wait times as a result of extensive IVR selections,” Shrivastav added. “AI can help this experience by accurately predicting wait times for a caller, so that they can make the right choice upfront about their ability to stay on hold.”

She pointed out that AI helps provide a level of personalization and empathy that each customer interaction deserves by predicting the sentiment and response to best balance empathy and outcomes.

“People are looking for more natural-feeling, human-feeling experiences, and there’s room for improvement here,” Bradley added, pointing out that many consumers feel like their experiences with chatbots have “lacked a human touch.”

He said this underscores the need for brands to consider what makes a conversation human — elements like empathy and understanding. “This is one of the reasons why tools that make it easy…to make bots more human are so important — they let non-technical experts like call center agents design and optimize bots, bringing all they know about good customer conversations into the mix.”

Bradley pointed out that humans don’t show up with a script and railroad you through it: they listen, understand and adapt.

Chris Hausler, director of machine learning at Zendesk, said emotional intelligence AI is important because it allows an effective means for dealing with people — many of whom are experiencing an issue.

“AI bots, like support agents,” he explained, “need to be able to provide empathetic and responsive service to help customers get to the outcomes they need — not just more quickly, but in a way that fosters goodwill with customers and keeps them coming back time and again.”

AI Tied to NLU Provides Rich Conversations 

To provide great customer service, conversational AI must incorporate Natural Language Understanding (NLU), which turns language into meaningful structured information for dialogue.

“A robust and accurate NLU is absolutely necessary for agents and chatbots to provide rich and flexible experiences,” Bradley said.

Shrivastav added that “AI can analyze hours of voice data to identify patterns in representatives’ performances and help managers improve onboarding processes and develop top performers faster. Customer support leaders can leverage insights of their teams to better understand the gaps and use those for training purposes.”

Bradley said that AI chatbots, similarly, can scale productivity without sacrificing customer experience.

“Instead of having employees handle routine inquiries like FAQs, order status or account support, AI chatbots can own these conversations, leaving employees free for more complex queries that are better uses of their time,” he said. “This is a win-win for employee productivity and customer experience.”

AI Can Raise the Bar for Data Analytics

AI has varied uses for raising the level and quality of analytic insights gathered from call center data.

When an agent or bot can access a customer’s history and data in one place, they can make conversations feel more helpful and personalized.

“By having a unified view of analytics and insights from all of their voice and messaging conversations, they can work on improving experiences across the entire customer journey,” Bradley said.

With AI’s help to measure consumer intent, brands can gather critical insights from verbatim messaging data to improve and optimize business operations. Call centers can also use AI insights to identify intents ripe for automation, like establishing FAQs, training agents on the weakest CSAT areas or tuning bots and agent scripts.

Shrivastav explained that HubSpot’s conversational AI allows users to capture voice data in their CRM, enabling them to unlock coaching opportunities, quantify competitive trends, surface top objections and identify changes in market dynamics.

“Customer conversations are such a rich source of actionable insights — the ability to track trends, outliers and insights from calls is a massive opportunity for call analytics,” she said.

Although these capabilities have existed in systems for a while now, she said modern AI can help explain the “why” behind unexpected scenarios by quantifying competitive trends and identifying changes in market dynamics. For example, a sudden drop in call volume could be a factor of product updates or external conditions such as bad weather.

“Understanding these reasons is critical for growth, and AI helps us get there,” Shrivastav said.

Hausler explained that by providing detailed insights into the customer’s intent, sentiment and language, AI can also improve call analytics.

“While your agents concentrate on giving each customer a great, personal experience, AI can monitor analytics that help inform training, team capacity, recurring issues or high-selling items and more,” he said.

He added that the benefits to agents aren’t limited to time-saving or improved analytics. “While AI takes the strain out of the more administrative elements of the job, agents can focus instead on providing the human touch to each interaction — a key element in fostering loyalty.”