The Future CX Skill Set Is Systems Thinking, Orchestration, and Knowing Where Humans Create the Most Value

The conversation around AI and customer experience is running ahead of most organizations' actual operating reality. Headlines focus on layoffs, agent-to-engineer ratios, and the promise of full automation, but for the majority of CX teams, the daily experience of AI is a basic Copilot license, a chatbot experiment, or a handful of generative tools used in isolation. The gap between what the market is discussing and what most teams have actually deployed is enormous, and it's creating anxiety without clarity. Leaders know they need to act. They don't yet know what acting well looks like because most of them haven't felt what collaborative AI workflows can do to their own work.

Robin Wong is the Vice President of Customer Experience at Bounteous, an AI services firm, where he helps enterprise clients reimagine customer and employee experiences through design-led, AI-enabled transformation. He previously spent seven years leading design and digital-first transformation at BT, and his background spans experience strategy, design thinking, and organizational change across financial services, telecoms, FMCG and tech. His vantage point across Bounteous' client portfolio gives him visibility into the full spectrum of AI maturity, from companies still in early experimentation to those two or three years into aggressive AI enablement.

"I think most leaders struggle to understand the power of AI and the positive change it can enable because they haven't felt what it’s like to actually reimagine how they work," Wong says. "Until you have one of those moments where you feel what it's like, it's very difficult to describe or even imagine the future." That experiential gap, which Wong is clear to distinguish from a technology gap, is what he sees holding most CX organizations back from making smart decisions about where AI should and shouldn't operate.

The maturity gap is wider than the headlines suggest

Across Bounteous' client base, Wong describes a spectrum. Very few are AI-averse. Everyone, he says, feels the pressure to move. But most are still in early experimentation, using tools at a surface level without integrating them into collaborative workflows or shared team contexts. The companies two or three years into serious AI adoption are talking about productivity and revenue-per-person rather than headcount cuts, but they represent a small minority. "The more mature players are focused on achieving business outcomes rather than just the blunt tool of headcount reduction, but I think the layoffs haven't really hit yet for most companies. The concern and anxiety are high, but the actual impact is felt less," he says.

He sees a meaningful difference between using AI as a tool and using AI collaboratively. At Bounteous, teams work on shared projects within Claude, Cursor, Codex or other AI platforms, using shared skills libraries, context, and memory. That kind of integrated workflow is what dramatically accelerates output. But when he describes it to clients, it feels distant from their current reality. "It's like a training session where someone tries to explain a concept, but when you actually do it yourself, you're like, 'Oh, I get it now,'" Wong shares. "Doing it collaboratively with other people is where the real power is."

AI can accelerate in places, but it can't deliver end-to-end CX

Wong is specific about where AI creates genuine value in CX today: synthesizing research into themes, generating rapid prototypes, personalizing offers and recommendations, pre-populating customer journeys, surfacing next-best actions, and connecting data across fragmented systems. These are real contributions that reduce friction and unlock revenue. What AI cannot do, he asserts, is replace the full scope of experience delivery. CX work depends on empathy, contextual judgment, complex multi-step thinking, and the ability to ask 'why is this happening?' five layers deep to understand what a customer actually needs. This requires context AI does not have access to.

Wong uses a set of recent experiences to illustrate this. He tried to get AI to produce a coherent slide deck and found the effort-to-quality ratio made it impractical. He watched financial services agents announced by major AI labs and concluded they do parts of what a junior analyst does, but not the integrated reasoning the role requires. "Gen AI is there to work out the next-best word or do pattern matching. It physically cannot empathize or have enough context to truly understand why a customer is saying something. It can't do multi-step design thinking. Human beings are just naturally more collectively intelligent in terms of the set of skills and capabilities required."

Over-optimizing for efficiency creates unintended consequences

Wong's sharpest warning is for organizations that treat AI primarily as a cost lever. When teams spend less time on the work they enjoy and more time managing agents and spinning plates, job satisfaction drops. When the quality of human interaction decreases because the human is stretched across too many automated workflows, the customer feels it. "There's a very joyless version of the future where you become an orchestrator of loads of agents and the time you spend doing the thing you love gets squished further and further down. As a designer, surely we can intentionally design something better."

He advocates for a sustainability lens alongside the standard desirable-viable-feasible framework. Is the AI deployment good for people, planet, and profit? Or is it technically feasible and commercially attractive but corrosive to the human experience it's supposed to improve? "It's easy to sleepwalk into doing something just because it's desirable, viable, and technically feasible," Wong says. "It's not always a good thing when you zoom out."

The CX skill set is shifting

For CX professionals looking at where to invest in their own development, Wong identifies several capabilities that are becoming essential. Curiosity and willingness to experiment across different AI platforms is table stakes. He sees a deeper shift toward systems thinking, understanding how agents, tools, data, and human workflows connect into a coherent operating model with shared context and memory. "Businesses need a system where all this stuff joins up," he emphasizes. "The client shouldn't have a different experience with agents than they have with humans. Having that systematic approach to operationalizing workflows is really important."

In his view, success in this area depends on a framework of four emerging roles: experimenters who push the boundaries of what's possible, operators who string the pieces together into reliable systems, professionals who ensure quality and explain decisions to stakeholders, and orchestrators who manage the interplay between human and AI capabilities. "It requires a more systemic view of what's going on," he says. "There definitely need to be people who are skilled in translating how these things could and should be reimagined in the future."