Generative AI: A new catalyst in customer support

Generative AI is already projected to generate over $400 billion annually in business value within a few years. One area where it’s already showing value is in customer service and support. 

Colin Crowley, a CX expert and former senior director of customer marketing at Freshworks, has an up-close view of how (and why) companies are leaning into generative AI to drive results. In an interview with The Works, Crowley identified customer support leaders’ biggest challenges and how generative AI is addressing those. 

Conversation edited and condensed for clarity.

What are customer service leaders’ biggest challenges right now? 

One paramount issue is the ability to increase customer service quality while simultaneously increasing efficiency. Historically, those two things were at loggerheads, and you were typically doing either one or the other.

You had efficiency-based metrics that prioritized speed and getting the customer off the phone. Or the priority was to get that email out, and quality suffered as a result. Now, AI and automation technologies are making it possible to increase quality and efficiency at the same time—and that's what customers increasingly expect. 

Another concern is that leaders are more conscious about making better use of people. How can I use the people in my team better? How can I optimize the use of their human skills of empathy and otherwise use machines to handle tasks that can be automated? How can these technologies improve agents’ experience and make it easier to retain great talent? 

What are the specific impacts of generative AI in customer support?

Generative AI is opening up access to AI. In the early days, if you wanted anything AI-related, you would have to work with third-party companies and integrate with your existing software. That made it difficult to experiment with AI if your company was not very supportive or did not provide a safe space or the budget to experiment. 

Generative AI changes that. Using open-source material, you can automatically create features such as our Rephraser and Tone Enhancer, where you can return results for how you may rephrase a sentence in real time, or how you may change the syntax or tone. The fact that you can just turn something on and receive value from that is new for a lot of companies used to waiting for a model to be built up and having a delayed ROI. Generative AI delivers more immediate value.

It also does a lot when it comes to data mining. Customer experience leaders are on top of tons of data. Their traditional problem has been multiple contacts and a lot of back-and-forth with customers. There’s a tremendous opportunity to mine that data and understand more from real conversations as to what phrasing or tonality works better for customers.

Getting that data and organizing it is a big challenge for CX leaders. Generative AI can help them get to that data and make sense of it in useful ways. There will be a lot of opportunities to utilize gen AI to empower employees and agents to deliver higher levels of customer service and be faster. 

How are AI-powered chatbots driving results in customer conversations? 

AI allows chatbots to do a lot more than they were able to do previously. Traditionally, people thought of personalization in sales-related contacts. But other aspects of personalization are increasingly intimate. One example is customer archetypes.

Every company has some way of segmenting customers, but AI provides opportunities to segment customers in much more interesting ways: number of seats, loyalty tiers, how long has someone been a customer, etc. You can also include data on how they engaged with you on social media and create more refined segments—and use those segments to provide personalized support. Personalized recommendations can help companies figure out, for instance, how much discount to give back to a customer in case of a service issue. Companies are trying to balance a good customer experience with how much money they are spending on that experience, and AI can help them find the sweet spot. 

How important is the constant training of chatbots? 

That's one area where companies still make a lot of mistakes. Even though the bot can get better at training itself, you need someone who is going to pay attention to how customers are interacting with the bot. You also need people who can ensure quality auditing of AI-related processes and regular metrics review so that you get the ROI you expect. A lot of companies don't invest enough in that regular maintenance of AI and automation.

How should CX leaders strike the right balance between automation and the human touch? 

The general truism is that companies are moving away from being at either end of the spectrum. Some have a “bots for the sake of bots” mentality: Chatbots are fancy, let's invest in chatbots! And then you have the other end: companies that think it's always better to have humans to answer customer queries. 

AI and chatbots are typically better used in circumstances where you have an issue that is repeated in high enough volume to justify investment. Chatbots provide the greatest value on things that are relatively straightforward, where they only need a customer to answer one or two questions to get to what the customer needs.

Other use cases include when something is not likely to be changed a lot—say, a policy or process—or where customers have a clear end goal in mind and typically don't need a lot of empathy because it's a pretty straightforward situation—an order cancellation, for instance.

But there is still an X factor companies don't necessarily know about until they do an A/B test or something. Companies need to discover what those areas of ambiguity are for them and strategically test those.

We want to hear from you! Please send us your feedback, and get informed about exciting updates from The Works. Drop us a line: theworks@freshworks.com.