Talk of ChatGPT is everywhere from your LinkedIn feed to the evening news to, yes, this blog. For many, the buzz about ChatGPT is evoking the same exploratory wonder they once experienced when the Internet itself first debuted. Like most people in the customer care sector, we are diving deep into where ChatGPT fits, or might fit, into the contact center world. The question of the moment is: what role is ChatGPT going to play in delivering a loyalty-building customer experience? So let’s get into it.
We’ve long been proponents of making transactional customer service low effort, for both customer and contact center. Value comes from focusing energy and resources on complex customer care scenarios that require human insight and human decision-making. This is the underlying philosophy of our digital customer experience support solutions, driven by AI, machine learning, chatbots, and intelligent self-service. But with ChatGPT in the spotlight, we wanted to bring you a deeper dive into AI in the contact center.
If you ask ChatGPT directly, it’ll tell you that as a generative AI-based chatbot, it can enhance the customer experience through its 24/7 availability, consistency, scalability, and personalization. Those first three are what we’d expect. They’re exactly what contact center technology has been trying to achieve for years, through IVR and chatbots and other automations that promise to increase the efficiency of delivering customer service. ChatGPT and other AI-based models are essentially a step up from the technologies many contact centers have used to support their simplest customer service transactions for years.
That fourth enhancement though—personalization—is the interesting one. If you’ve taken the opportunity to play around in ChatGPT, peppering it with increasingly conversational questions, then you’ve seen the power of natural language processing at play. There’s no question that it can mimic a human, taking on a natural and engaging tone to interact with customers—especially in those more transactional instances.
We’ve all experienced the particular hell that is being stuck in an endless IVR loop where you end up stabbing ‘0’ to get to a human, so we can say with assurance that customers will appreciate the sophistication ChatGPT could bring to automated customer service. It’s more intuitive and versatile, with infinite applications outside of the usual “if/then” conditions that call centers have been tied to in the past.
One survey of consumers who frequently interact with customer service bots reveals that 77% believe AI is helpful for simple issues. They also are convinced AI will improve customer service efficiency and can match the service level of human agents.
It’s obvious that the biggest bang from ChatGPT-style AI tools will be improving both the quality and efficiency of transactional customer support, but what about the realm of sophisticated customer care we’ve historically reserved for more highly trained human agents?
“I can analyze language and tone to infer a customer’s emotional state,” ChatGPT told us; but “it is not a perfect system and there may be limitations… Emotions can be complex and nuanced, and there may be cultural and contextual factors that can impact interpretation of language and tone.” It conceded that empathy and situational decision-making—two skills that are critical for complex customer care—are areas AI would struggle to achieve. Data and algorithms can only do so much.
However, AI tools like ChatGPT can also serve in the capacity of a virtual assistant to frontline contact center agents. A customer calls, and it can instantly parse all historical data in correlation with the new request, prompting the agent with a particular approach or solution. It not only saves the agent’s time, but it can also help ensure a consistent response. In this way, the tools will complement human skills, provide real-time support to agents, thereby improving the agent experience and the customer experience concurrently.
64% of leaders expect that AI will inevitably replace some jobs—its efficiency and scalability ensure big cost savings over time. McKinsey estimates 60% of occupations could be automated over the next few decades. However, many experts optimistically state that it will lead not so much to a replacement of people, but a reframing and restructuring of what their jobs entail.
This especially resonates in the contact center world, where high-volume-high-stress-low-complexity transactional customer support, particularly in scenarios where customers are upset and irate, typically results in high rates of employee burn-out. AI’s potential to improve automated handling of these transactional interactions with better intuition and nuance could change everything. It would mean contact center employees will be deployed only in situations where their human-ness, their skills in decision-making and building true connection, are required and valued.
From that perspective, there is a case to be made that the strategic implementation of AI will also result in secondary improvements beyond typical contact handling metrics like AHT—perhaps lowering agent attrition, for example. The magic combination for retention is: meaningful work + support + empowerment = engagement. If ChatGPT can help an agent feel supported and better positioned to achieve resolution for an escalated customer, all while quickly and effectively diffusing tense situations, we’re betting you’re going to see less agent burn-out on your toughest programs.
As more contact center tasks become effectively automated, the need for personalization is paramount to protecting a company’s brand and customer relationships. And that’s one of the biggest benefits of ChatGPT-style AI.
With access to data such as purchase history, browsing behavior, interaction history, demographic data, and more, ChatGPT’s AI language model would be adept at personalizing the customer experience—in addition to boosting efficiency, responsiveness, and consistency. It’s an opportunity to lay to rest the dreaded robotic response that says: “Sorry, I did not understand your question. Let me transfer you to an agent.”
But true personalization is dependent entirely on an extensive cache of data. AI tools must be trained with relevant, highly tested datasets from across the company’s systems. Contact centers that consider implementing advanced AI solutions need to be equally invested in their data strategy.
However, many organizations are deeply aware that their data is siloed, and they don’t necessarily have the resources to resolve this challenge. There’s also the inevitable questions that arise surrounding data privacy and storage—something contact centers are already quite familiar with.
Flawed data isn’t the only risk. So far, many companies admit their investments in AI have been ad hoc rather than strategic. For those that are investing in AI as a solution specifically to automate tasks and cut costs in the process, short-term achievements will likely outweigh long-term success, according to Forrester. The risks can quickly add up. Maybe it will be inadequate data inputs or unexpected customer backlash—organizations need to be prepared for the big-picture impact of AI in customer care.
One criticism published in the Wall Street Journal recently brings up some valid points: In one example of a contact center using a ChatGPT-style bot as a virtual assistant, the atmosphere is fraught. It reportedly feels like a micromanager who doesn’t have the whole story; errors abound, and agents’ performance is poorly judged. Though initial results look positive—reduced average handle times and faster resolution—there were other more subjective issues that weren’t anticipated, such as higher agent stress as they struggled to adjust to this hands-on AI tool. Despite advanced sentiment analysis and natural language processing, there are simply some scenarios—even seemingly transactional ones—where human connection is the better strategy for long-term customer loyalty.
All in all, there is no question we will see a transformational shift in customer experience design over the next few years, with AI/ChatGPT tools driving rapid change. It has the power to relieve the burden of transactional customer service while still wowing your customers and staying true to your brand—but only if it’s strategically implemented with a view of the big picture.