At Customer Contact Week’s Innovation Summit 2026 in Las Vegas, I joined a panel to unpack CMP's Annual Consumer Preferences Survey¹, a look at how consumers want to engage and what they expect from AI-powered customer experiences, including how they feel when interacting with a chatbot or AI agent, what they request when escalating to a human, and how they’re using AI to augment their own communication.
The survey itself told an encouraging story: 45% of customers say their experience got better this year, compared to 37% who say it got worse, a sharp swing from last year, when 59% reported things getting worse. Progress, yes. But not the full picture. Only 38% of customers find their service experience convenient. Just a quarter feel it's personalized or feel they have any say in how an interaction happens. And only 18% say the experience is consistent across every channel they use.
What I shared is that customers leverage all channels that they are in. What matters is that the problem gets solved, and that the handoff from AI to a human happens in the moment it's needed, orchestrated in real time, not three menus later. But that indifference has a catch, it's a learned behavior. A bad first touch with a bot or other self-service application teaches customers to avoid them next time. Deploy AI without first process mapping the flows and getting the underlying data and infrastructure right, and you're not neutral, you're actively making the experience worse. Getting the data and process right before you build on top of it isn't a preliminary step; it is the essential foundation to create a seamless experience.
That's the piece I don't think our industry has solved yet: not whether to use AI, but where the data feeding it actually originates, and whether it's layered onto processes and infrastructure that were never optimized in the first place. The unglamorous work, cleaning up the data layer, optimizing the IT foundation underneath, is what decides whether anything you put on top of it performs.
In designing, deploying, and managing these AI-enabled channels, the conversation turned to how we are enabling our agents. In my experience working with clients, agents are heading into the best career moment they've had, not despite AI, but because of it. Take a simple example from banking: a customer calling in panicking over a possible fraud charge. An agent armed with full, real-time context, guidance surfaced as the conversation unfolds, recommendations offered before they're asked for, knowledge and tools one click away rather than five, can resolve that moment with confidence instead of suspicion. The same data can flip a routine call into a genuine opportunity, surfacing a product or service the customer needs, at the moment they need it.
This perspective reframes what "skills for the future" means for frontline talent. I highlighted two that are especially important:
As products and services keep commoditizing in banking, insurance, retail (and beyond), experience becomes the thing that actually differentiates a brand and keeps customers engaged with it. While brands compete on experience, the contact center is on the front lines of delivering it. Which means the real strategic bet isn't the AI stack. It's the investment in both technology and the people / culture that make that experience real.
The customer of 2030 is already telling us what they want. The question I'll leave you with is whether we're building the foundation, and the talent, to deliver it. My own view is that the answer lies in human expertise and AI working together, not AI standing in for people. If that's the question on your desk too, I'd welcome the conversation.
1) CMP's Annual Consumer Preferences Survey: https://www.customermanagementpractice.com/cmp-events/ccw/las-vegas/schedule/