CCW Vegas Dispatch: AI in Customer Experience Needs Humans in the Loop

Takeaways from #CCWVegas 2025

Las Vegas, June 2025. On the main stage at Customer Contact Week, a Marriott executive echoed founder JW Marriott: “Take care of the associate, they’ll take care of the customer.”

In a week dominated by AI demos and automation promises, this simple truth stood out. Even in 2025’s AI (AGENTIC AI) powered customer journeys, people remain the core differentiator.

Reality Check: Mind the AI Gap

The AI buzz in customer contact is real—but so is the implementation gap. By 2025, 86% of customer service leaders say they’ve tested or implemented AI (Sobot.io). Yet only 25% have successfully scaled automation into daily operations (AmplifAI, 2025). That’s three out of four contact centers still stuck in pilot mode.

The message from CCW Vegas 2025: AI won’t fix CX unless paired with people. Companies succeeding are taking a blended approach—combining automation with human insight, empathy, and judgment.

Three Pillars of Orchestrated Journeys

Across sessions featuring leaders from Marriott, FedEx, Equitable Bank, and more, three principles kept surfacing:

1. Intent-Driven Automation

AI works best with clear, repetitive tasks. HumanN, a nutrition brand, built a virtual agent named Hannah to handle multilingual, low-effort chat intents. Hannah now resolves 65% of all chat interactions (Talkdesk, 2025).

Why did it work? Agents helped train Hannah, testing it in real scenarios. The tool became a teammate, not a threat. Automation stayed focused: it handled known intents fast and accurately, and escalated edge cases to humans.

2. Human-First Hand-offs

When AI hits its limits—or when empathy matters—humans need to step in seamlessly. Career Certified equipped reps with AI copilot tools to generate real-time answers and call summaries. The result? 30-second drops in handle time without job losses. (Talkdesk, 2025).

Good design means AI defers to humans at the right moment, with context passed along. Marriott’s Frid Edmond underscored that some problems require observation, not dashboards: “AI today couldn’t have told us that. We had to see it, touch it, feel it.”

3. Continuous Learning Loops

FedEx's Neil Gibson emphasized that AI rollouts don’t end at launch. His team uses the ABLE framework (Assess, Build, Launch, Execute)—with “Execute” meaning continuous iteration, not just deployment.

Equitable Bank’s leaders echoed that sentiment: AI isn’t “set and forget.” They conduct regular quality audits and feedback loops. One company had to quickly retrain its bot after it mistakenly tried to upsell a customer. The lesson: governance and oversight matter.

Emerging Playbooks for AI + Human CX

Here are five moves top teams are already making:

• Start with real problems. Don’t chase shiny objects—focus AI on high-friction or high-volume use cases.

• Co-create with employees. Involve agents early to shape tone, test logic, and flag gaps.

• Define guardrails. Be explicit about what AI can’t do—set ethical and operational limits.

• Design seamless hand-offs. Ensure escalations are smooth, with history and sentiment passed to the human.

• Stay iterative. Schedule regular tuning sessions; retrain bots, update knowledge bases, and review transcripts.

Hard Questions Leaders Should Ask

When adding AI to the journey, smart leaders pause to ask:

1. What customer or employee pain point are we solving?

2. How does this tech help our people shine, not just cut costs?

3. Where do we insist on human judgment?

4. What’s our plan for feedback, governance, and accountability?

5. Are we using AI in a way that reflects our brand and values?

A Quick-Start Framework: ABLE

FedEx’s approach to AI implementation uses four phases:

1. Assess – Find a repeatable task or bottleneck to improve.

2. Build – Configure your AI tool with employee input.

3. Launch – Pilot in a limited environment. Measure outcomes.

4. Execute – Scale with ongoing oversight and refinement.

This model helps avoid “AI theater” and builds real capability over time.

Final Takeaway

Zack Kass of OpenAI called this the “Next Renaissance” for work. But the companies that stood out at CCW didn’t chase hype. They showed discipline—investing in their people while exploring tech or AI.

My takeaway: AI in CX is here to stay. But your humans? They’re here to lead.

It’s not automation vs. empathy. It’s orchestration.

And we’re just getting started.

Let me know how your team is blending AI and human service. Drop a comment—what’s working, and what’s still a work in progress?

I’d love to hear from you. What’s working (or not) in your AI + human journey? Drop a comment.

Stay Thoughtful - Guneet

#CCWVegas #cx #ExperienceMatters

Disclaimer: The views expressed in this newsletter are solely mine. I am not a spokesperson for my employer, nor do I represent my employer's opinion.