The smartest bets on AI for customer service

Our roundtable of leading experts offer their insights on a high-stakes business challenge

Blog
Jeff Davis

Jeff DavisEditor in Chief at Freshworks

Apr 04, 20246 MINS READ

Improving customer experience (CX) is an urgent priority for every organization today. 

That’s one reason why the market for CX technology and tools is worth $17 billion today and growing rapidly. The array of services available to digitize and personalize customer experience—including many now powered by AI—allows service leaders to implement critical changes faster and more directly than ever. 

The challenge many leaders face with AI opportunities isn’t so much about what or why they should implement, but where to commit resources across the spectrum of new CX strategies and technologies—whether it’s investing in better generative-AI-powered chatbots, beefing up data infrastructure to support AI, or deploying AI to empower and improve their best human agents, not to get rid of them.

Realizing there are no magic-bullet answers to those questions, we asked a handful of customer service and technology leaders what CX strategies they would prioritize in 2024. 

Build customer trust in your AI

— Mukesh Mirchandani, SVP of Global Solution Engineering, Freshworks

Blog

Many businesses experimented with AI in 2023. This year, they’re shifting from the experimentation phase to investment and execution. 

As they move along in their journey, they need to consider the issue of faith in AI.

Before the advent of generative AI, AI models sometimes failed accuracy tests. Some of that was due to the data that was available to train these models. But with gen AI, training data for large language models (LLMs) is no longer an issue. So how can we ensure that LLMs are using the right learning data? How can we know that the answer generated is relevant to our business and that customers can trust their responses? 

This lack of faith comes not just from occasionally errant results, but from a lack of understanding of how AI models work. Even if I tried, I would have a hard time explaining how LLMs work. I might convey some basic ideas, but that won’t build trust. A sales team might be more likely to trust its instincts over those of an AI tool that doesn’t disclose its methodology. In short, businesses need to commit to the “explainability” of the AI they use. 

Explainability is about AI “showing its work” to end users—to describe how it is making a recommendation, answering a question, or providing an insight. When a customer asks an AI chatbot about returning a particular product, how can the customer be sure that the chatbot has the most pertinent data on the company’s return policy? 

How, then, should companies make their AI explainable? Federal or state regulation could soon become a forcing mechanism, but as Freshworks President Dennis Woodside suggested recently, business leaders face a simpler question: “Are you ready to embrace AI transparency on your own terms for customers and employees now, or plan to wait to be told how to do it?”

The answer is obvious: Make AI explainability a priority now, not later.

Mine your customer conversation data

— Bruce Temkin, head of XM Institute, Qualtrics

Blog

One of the most underused assets companies have is the content from their interactions with customers. Organizations with maturing CX programs that have contact centers need to start analyzing their customer conversations.

Gen AI in particular is proving to be disruptive in data summarization, pattern recognition, conversational interface, and content personalization. I expect there will be a big push toward using these capabilities to handle more unstructured data sources, which will eventually lead to a new wave of CX adoption as the technology turns this data into much more actionable and accessible insights.

It's like two waves are merging to build a tsunami: the increasing use of unstructured data sources like contact center conversations combined with the power of generative AI to process and interact with all types of data.

Every vendor will be experimenting with where and how to use gen AI. Most of the year will be spent fine-tuning their approaches, so I would expect vendors to roll out a lot of truly impactful capabilities in 2025 and beyond. Combined with valuable customer conversation data, this will create an ongoing flow of insights that will transform how contact centers operate.

Subscribe to The Works

Insights on the business impact and ROI of AI. Sign up today.

Get AI into the hands of your service agents

— Shep Hyken, longtime CX researcher and author

Blog

We need to start thinking more about agent support with AI. We’ve spent a lot of time thinking about using AI with customer support—allowing customers to get basic information quickly, freeing up the agent to focus on customers with complicated questions. That creates a better agent experience in itself, but that is just one opportunity. 

There’s another opportunity to lower the average handle time, which goes up when the questions get harder. Agents no longer need to know everything. They just need to know how to find everything.

We can use AI to support agents' questions. When a customer calls with a complicated question, an agent can use AI to find the answer and communicate it to the customer in a way that suits them. If they're upset, calm them down; if they seem concerned, make them feel confident—showing empathy and creating a connection with the customer. Not only does the customer receive better care, but agents are far more fulfilled.

Making agents feel fulfilled could be more important than anything. When an agent is fulfilled, it not only means they are more apt to stay, but it also means the training costs involved in hiring a new person are eliminated. That’s a big incentive for leadership. 

Related: Shep Hyken on why agents need AI more than customers

Fully empower and equip your CX teams

— Mary Piercy, vice president of customer success, Simbe Robotics

Blog

I see an opportunity for increased connectivity supported by AI, especially as the role of CX grows in complexity. Client experience teams are becoming full-fledged participants in go-to-market strategies. It’s an opportunity to deliver a more seamless experience from the first touchpoint through the entire customer lifecycle. This early involvement empowers executive teams with a more comprehensive view of the customer's experience from day one.

Even still, companies don't have the luxury to invest in teams that don't move the needle on their financial priorities. CX teams need to show their impact in real dollars—both from a growth share and renewal perspective—and we need tooling that is increasingly supportive of that.

Mobilize and centralize your customer data for AI

— Annette Franz, CEO, CX Journey

Blog

Data is at the heart of designing and delivering a great customer experience. To make that a reality, though, we need the right tools.

There are so many disparate data sources and legacy systems. Platforms all need to be integrated. We can’t have tens or hundreds of systems and platforms that we're working from. There needs to be a central platform, and AI is going to help bring that all together.

AI is the best tool we have to derive insights and get those insights to the right people at the right time. It’s going to continue to be incorporated into every aspect of the CX tech stack.

Learn how leading companies are leveraging AI in support of CX. Download the free report (registration required).