LAS VEGAS - At Customer Contact Week Las Vegas, CX leaders from Walmart, Fanatics, and Wellby Financial shared hard-won lessons on deploying AI in contact centers. Their message: AI is a powerful multiplier, but it can't fix broken processes, and the most visible use cases aren't always the ones that deliver the quickest returns.
AI is a tool - not a solution
Bob Sacunas, VP and strategic industry executive at UiPath, used a blunt analogy. "You could drive in a nail with a wrench in a pinch, but why would you do that if you have a hammer?" he said. "And yet, I see a lot of organizations doing the same thing with AI. We're under pressure to do more with AI, and we start throwing it at things where it's really not the best solution."
The conversation at Customer Contact Week offered a practical view of AI for Customer Support, a field where hype often outpaces results. Anderson Wilkins, director of product management, agentic self-service and AI defect detection at Walmart, warned that AI is not a fix for fragmented customer support. "If your policies are inconsistent, your data is incomplete, if your channels give different answers - AI is not going to magically fix that," Wilkins said. "AI is a multiplier. It's going to multiply really great experiences, and it unfortunately can multiply some really poor experiences for your customers and members."
Transparency requires more than a disclosure
Simply telling customers they are talking to a bot is not enough, according to Megan Merrick, director of collector experience at Fanatics. "Is that experience working well? Then yes, you're going to build trust with your customer base, they're going to feel better about reaching out to you. But if it doesn't work, you're losing customer trust to a higher degree."
Wilkins added that transparency often gets buried in legal jargon. He argued that disclosures should explain what the company is doing with AI and why, in language customers can understand. Merrick designs AI interactions with the failure state in mind, focusing on minimizing frustration when things go wrong rather than just chasing the maximum potential.
Start with tools that support your team, not just your customers
Customer-facing chatbots grab attention, but the panelists emphasized that internal AI tools often deliver faster, clearer returns. Brent Nelson, VP of the virtual communication center and virtual experience at Wellby Financial, described a shift in strategy. "We went from zero to 100 trying to appease our membership, when in reality we should have started with our team members, because then we would have seen a much better result and a much quicker return on our investment."
Wilkins agreed, pointing to AI's ability to analyze contact drivers and repeat calls. "I think the full goal of this is utilizing AI to understand what are those contact drivers, what are those intents, what is the friction behind those repeat calls," he said. "To be able to synthesize the data and be able to point to that information, that intelligence, at exactly the right processes and products that actually need to be changed."
Why this matters for customer support professionals
The panel's advice boils down to three concrete steps: audit your own processes before deploying customer-facing AI, design AI interactions with the failure state in mind, and invest in tools that help agents and supervisors before rolling out chatbots to customers. As Wilkins put it, the ultimate goal is to use AI so effectively that customers never need to call.
Your membership also unlocks:
