Artificial intelligence isn’t going anywhere. In fact, it’s beginning to take the lead on customer interactions — but not without challenges.
Years ago, interacting with a branded bot was a pain for customers. It meant pressing numbers on a telephone keypad (or hitting 0 repeatedly) but not getting the right answers. Typing questions into a website chat box while trying desperately to find a customer service phone number (and some companies went above and beyond in trying to hide that number).
Those interactions often ended in frustration, anger and even jumping ship to another company willing to put a human voice on the line.
But those days are gone, and people’s perceptions of bots have shifted massively. In a survey of 6,000 people from Pega, 55% of consumers said they’re comfortable with a business using artificial intelligence to interact with them — and 26% were neutral on the topic.
Additionally, 68% of those surveyed said they’d be more open to AI if it helped improve their daily life in some way — such as by saving them time or money.
Today, brands are using more than chatbots and “dumb” phone bots to engage with customers. Artificial intelligence offers up a whole host of innovative opportunities, both on the customer and employee side. This technology can solve customer problems faster, free up worker time, unify data across the organization and so much more.
“AI tools will be shaping the future of customer service — this seems to be the way the world is heading, with technology becoming more and more integrated into business tactics,” said Sourabh Gupta, co-founder and CEO of Skit.ai.
Let’s take a look at those tools, how brands are actively using them and some of the challenges.
Conversational AI Powers Human-Like Conversations
Let’s start with conversational AI, because this technology will feed into a lot of other applications we’re going to cover.
Conversational AI is exactly what it sounds like — a computer program that can interact in a human-like way with consumers. It often takes the form of chatbots, a voicebot that you interact with on the phone or even a personal assistant, like Amazon Alexa or Apple’s Siri.
According to Dinesh Nirmal, general manager, IBM Data, AI and Automation, “Conversational AI, at its core, is a solution designed for businesses so they can better serve consumers.”
Nirmal emphasized that this technology is ultimately designed for a person trying to access information, solve a problem, pay a bill or schedule an appointment. He pointed out the example of CVS Health using this technology during the COVID-19 pandemic.
The health retailer began using the IBM Watson Assistant to streamline its telephone system while rolling out the COVID vaccine. IBM Watson could respond to questions about eligibility, side effects, needed documents, cost and more.
The Challenge With Conversational AI
While the benefits of conversational artificial intelligence are clear, the technology is not without a learning curve.
“At its core, conversational AI software must perform three things: understand the user’s question; find the best answer from its training data or search for it through documents/content; and return an answer in a concise, precise manner,” said Nirmal.
That first part of the equation — understanding the initial question — is essential. “And this is where the messiness and complexity of human communication present the greatest challenge,” explained Nirmal. “People don’t always use the same words to describe what they need, they speak different languages, and they frequently get confused or use the wrong terms.”
One of the benefits of conversational AI with natural language processing (NLP), however, is that the technology can continually learn and improve over time. The more customer interactions it handles, the smarter and savvier it becomes.
But a few things brands must keep in mind, according to Nirmal? “Conversational AI systems need to be designed with the end-user in mind, able to integrate with many different back-end systems and easily integrate with your contact center system so human-agent hand-off is seamless.”
Related Article: 5 Best Practices for Using AI in Your CX Strategy
Chatbots Get Smarter, Work Harder
The chatbot has been a staple in brand experiences for years — and it’s only getting smarter, powered with the technology we covered above, conversational AI.
Virgil Wong, chief data officer at HGS, told CMSWire AI customers want well-designed chatbot interactions — especially ones that can offer quick solutions to problems. But to ensure this design, businesses need real-time, actionable customer analytics to enhance bot responses, Wong explained.
He also added that “sentiment analysis is critical to determining the next best action, getting the correct answer fast, detecting fraud or compliance issues and summarizing interactions for follow-ups and reporting.”
The Challenge With Chatbots
While today’s chatbots are getting smarter, they still have their limitations, and humans will still be necessary to shoulder the more complex or creative challenges that arise.
According to Wong, some of the biggest issues brands come up against with this technology include strategic considerations, such as how much to invest, along with technical concerns, like bugs and program errors.
“It’s essential to understand the strengths and limitations of chatbots — and how it fits into your larger strategy of optimizing every point of the customer experience,” said Wong. “Too many brands build half-baked bots without a clear roadmap, analytics strategy and concrete plans for iterative development cycles. Customers don’t need chatbots that serve only as barriers to connecting with a live human agent.”
To enhance your chatbot, Wong recommended:
- Using a skilled writer to tap into the creative side of engaging conversations
- Onboarding writers and programmers that can communicate using the brand’s voice while responding to customers’ tonality and emotional intent
- Hiring a CX partner that can help plan, implement measure and improve your chatbots
Voicebots Can Assist Customers 24/7
Voicebots are another technology powered by conversational AI. They can communicate with consumers verbally, typically over the phone. You’ve probably interacted with one yourself when you’ve called a company’s customer service hotline.
This technology comes with a lot of hard-hitting benefits. “With Voice AI, brands can cut down on customer wait times, shorten the time they spend on the phone, and effectively answer their questions faster,” said Skit.ai’s Gupta. These bots can also operate 24/7, something that’s necessary in today’s always-on world.
According to Gupta, voice AI can impact a lot of business processes, including sales procedures like outbound calls and proactive customer service. “The key is to identify these use cases that often arise out of a business’s specific needs,” he said.
And this tech doesn’t need to be robotic and impersonal — like the “press 1 for x, press 2 for y” prompts you may have heard in the past.
“Voice AI can be as personalized and as nuanced as a company chooses,” said Gupta, “and this depends on the way AI and ML models are trained, and the effort poured into annotating the emotional nuances.”
The Challenge With Voicebots
Similar to what we mentioned with conversational AI, a lot of the challenges associated with voicebots come down to verbal language — both what’s said and what’s not.
“The challenge for a digital voice agent is twofold; not only does it need to identify emotions, but it also needs to respond in real-time,” said Gupta. “Unfortunately, digital voice agents are not nearly as emphatic as human agents are.”
And if customers are not expressive, it can be hard for the technology to identify specific emotions. For instance, bots might have a problem picking up sarcasm cues or identifying an irate customer who expresses themselves through excessive silence.
But Ski.ai is attempting to overcome these voicebot challenges by delving deeper into their work on prosody, which looks at how people use stress and intonation within their speech patterns.
“When a person ends a sentence,” Gupta explained in an example, “their manner of speaking is different than when they continue to speak. The right identification of prosody will help the digital voice agent respond quickly, and make the conversation more engaging.”
AI Assistants Can Help Employees, Interact One-on-One
AI-backed agent assist technology can go beyond answering questions or scheduling appointments when it comes to making workers’ lives easier.
For instance, say an employee is on the phone with a customer who has a question, and that employee has an AI assistant at the ready. The AI can automatically interpret what the customer is asking, search a knowledge base for the answer, and display the information on the employee’s screen while they’re still on the call.
These assistants, powered by conversational AI, can also hold one-on-one conversations with customers themselves.
“Advanced AI Assistants…interpret buyer responses and respond accordingly, all on their own, until that contact self-identifies as ready to talk to sales or customer success for the next step,” said David Greenberg, Conversica CMO. “Every action is automatic, personalized and proactive … .”
Some one-on-one tasks an AI assistant can complete include:
- Connect with leads after an event
- Reach out to former customers to win them back
- Connect with unresponsive prospects
- Open renewal conversations early
The Challenge With AI Assistants
AI assistants, when interacting with customers, need to hold conversations that feel personal and humanlike, according to Greenberg. The tech also needs to be able to interpret intent and respond accordingly — all while keeping responses natural.
To do this, Greenberg said brands must look for AI assistants that work on multiple channels and have natural language processing and dynamic message generation capabilities.
“The best Conversational AI solutions will also be battle-tested, with a long history of real-world interactions that the AI can learn from and improve its accuracy,” he explained.
Related Article: Real-Time AI: A Necessity for Great Customer Experiences
Humans Still Necessary in an Artificial Intelligence-Filled World
Artificial intelligence tools offer up a lot of benefits — but ultimately, that human piece of the puzzle isn’t going anywhere. Not only are people necessary to fill in the gaps of AI’s limitations, but it’s something customers, at the end of the day, crave — especially when they have a problem they need solved.
“We believe the purpose of AI is to augment human intelligence, not replace it,” said Nirmal. “We believe AI is going to transform the way we all work for the better, free people from tedious, non-value-adding work and ultimately help create jobs around the world.”
Greenberg added that businesses that tap into these technologically advanced tools now will be better equipped to meet customer expectations, especially with a potential economic downturn on the horizon. “Their digitally-delayed counterparts,” he added, “will only be left further behind in a recession.”