It’s an innovation that brings great benefits. It’s a disruptive force that will displace millions. The emergence of artificial intelligence, whether through tools such as ChatGTP, the typing assistant Grammarly, or digital voice assistants, elicits varied and strong emotions, many of them centered on the potential effects it will have on jobs. For every perceived benefit, there is a corresponding warning, the latter usually centered not around the technology itself, but how it is used.
AI and machine learning are parts of customer support, too. Parts, not the sum total of the enterprise and in our industry, opinions on the subject are as varied as they are anywhere else, ranging from being seen as indispensable to being an innovation that causes more problems than it solves. Perhaps the question is not so much about the utility of AI but about how it is used.
Resolving the False Dilemma
Discussing the topic goes off the rails when AI is framed as part of a binary choice as if BPOs must choose one or the other. We are in the ‘both’ camp. Contact center operations are enhanced by technology, but they are powered by people. Support is a people-centric enterprise – a person with an issue or question connects with a person on the customer support team for help. If you’re thinking of AI as the replacement for live agents, you’re doing it wrong.
The quality of the overall experience is judged on response time, the agent’s comprehension of the issue, the accuracy of information, and so forth. As technology becomes more complex, so do the problems that customers encounter, and this is one area where AI can enhance efficiency. We use it to augment the live agent as a virtual assistant or coach. This prevents calls from being transferred with customers having to repeat themselves, two significant sources of frustration when connecting with the support team.
In a stand-alone capacity, AI is a valuable tool when self-help is a viable option for simple tasks and basic questions. The whole point of having FAQs lies in the name – these are the topics customers ask about most often, and they are usually uncomplicated. An automated system provides those answers or points customers toward a knowledge base or other resource. Both sides benefit; customers get quick access to needed information while the team’s workload is better managed so agents can spend more time on issues that require it.
Left Brain, Right Brain
AI excels in analytics, whether that means sifting through data, predictive ability, or instantly providing accurate customer records. The balance to that is human empathy, intuition, and picking up on customer sentiment to ask meaningful follow-up questions that get to the heart of an issue. The two combined provide a whole-brain support experience that can deliver personalized service in support of resource-intensive industries such as healthcare, hospitality, and education.
For instance, AI is an integral piece of our coaching portal, providing insights into agent behavior that are used in formulating coaching plans. In a service environment, the technology’s predictive ability is used to better understand customer sentiment, buying habits, and channel preferences, resulting in a more personalized customer support experience. While tools like ChatGTP are trained on existing human content, they cannot be trained to be human.
What is gained in efficiency and a memory that never forgets is lost in terms of a live agent’s ability to think in the moment and connect with a customer at the individual level. Agents are also valuable in collecting feedback about customer sentiment, again picking up on verbal cues that arise during an interaction. This feedback is an invaluable tool for understanding what customers think of a brand, what new features they would like to see, and information that companies can use in moving forward.
The Future Is Here
The introduction of AI presents a trend and a challenge. The trend is to use its analytical power and give agents a proactive tool for resolving customer issues. The challenge is in managing the technology by understanding what it is and is not. First, AI is not a replacement for agents. What it can do is make agents more productive and better able to tackle high-value customer concerns. Second, its analytical power can be used to “understand” customer sentiment and triage calls, chats, and emails accordingly, which helps to minimize escalations and prioritize responses. Third, it can be trained to spot customer habits and preferences based on their previous activity.
That last point is important because AI is only as good as the data used: is it consistent, accurate, or has enough volume to allow for meaningful analysis? Any shortcomings in the information put in will be reflected in what is kicked out. AI ‘learns’ from each data set that is entered and it does not forget. Instead, the knowledge base builds with each new data dump, and over time, predictive capability comes to the fore. Memorable experiences start with engaged agents; AI helps to make agents better.
