Editor's note: This article originally appeared on forbes.com. John Abou is a Forbes Business Council member.
There is a problem playing out in customer experience right now.
Enterprises are investing aggressively in AI, automation and analytics to transform how they serve customers. Yet many are still managing technology and service delivery as separate functions. Platforms are selected by one team. Operating models are managed by another. Frontline associates are trained downstream, after the technology decisions have already been made.
In today’s enterprise AI adoption, that model can be costly.
According to a recent Forbes article, research from CX platform Nextiva found that 92% of companies are in some stage of bringing AI into customer interactions, while 55% are still in the early stages of adoption. The same study found that nearly nine in 10 respondents believe the best customer experience will come from a balance of automation, AI and humans. However, a 2025 report found that 82% of consumers prefer human support, even if AI outcomes and wait times were the same.
That tension captures where the CX market is today. AI is today’s Wild West and there is a race to implement it within existing systems while many customers still have reservations or outright frustration about interacting with it.
How can companies build the operating model to make AI useful, trusted and effective in real moments that matter for customers?
The answer: True orchestration.
What true orchestration looks like
When technology teams and service teams operate independently, gaps appear quickly. A virtual agent may be trained on policies that do not reflect the way customers actually describe their problems. An associate assist tool may surface the right information, but at the wrong point in the conversation. A bot may resolve simple requests efficiently, but hand off complex interactions without context, forcing customers to repeat themselves.
None of those failures are technology failures alone. They are design failures. More specifically, they are orchestration failures.
AI alone does not create great customer experience. It needs operational context, clean data, governance, feedback loops and people who understand the emotional and practical realities of customer interactions. Companies that are more likely to stand out connect their digital intelligence with their operational intelligence, which has been informed by human experience.
That requires a different way of working.
In the past, a company might have treated the contact center as the place where a technology strategy was executed. Now, the contact center should be where that strategy is formed. Frontline data, associate feedback, customer sentiment, quality insights and operational patterns should inform how AI is trained, where automation is deployed and when human judgment should take over.
This is where many organizations need to rethink ownership. AI in customer experience cannot sit entirely inside IT, digital, operations or marketing. It cuts across all of them. That’s one reason more companies are creating dedicated AI leadership roles, AI councils or cross-functional governance groups to connect strategy, risk, data, workforce planning and execution. In IBM's 2025 survey of CEOs, 50% said rapid investment of AI had left their organizations with disconnected, piecemeal technology. The same study found that 68% viewed integrated enterprise-wide data architecture as critical for cross-functional collaboration.
For CX leaders, this has practical implications.
AI product owners work directly with operations leaders, not just technical teams. Quality teams move from sampling a small percentage of interactions to using AI-enabled analytics that can identify patterns across far more conversations. In my experience, AI-enabled approaches can analyze interaction patterns across channels and surface coaching, compliance and customer friction opportunities more quickly than traditional quality assurance tools.
Prepare for new roles. Have conversations with designers who understand both brand voice and operational policy. Put AI trainers and supervisors in place to evaluate outputs, identify failure patterns and refine knowledge sources, just to name a few.
Expect roles to evolve. Workforce teams will need to plan not only for staffing volumes, but also for new skills in judgment, empathy, exception handling and AI supervision. Operations leaders will need to become more fluent in data. Technology leaders will need to become more fluent in the realities of service delivery.
AI interactions built for customer intent and trust
This isn’t just a simple structural change. It’s a mindset change. The old approach was, “Which interactions can we automate?” The better question to ask now is, “What should each interaction accomplish, and what combination of AI and human expertise will produce the best outcome?”
Sometimes the answer will be full automation. A customer checking a balance, changing an address or confirming an appointment should not have to wait for a human being if AI can resolve the request accurately and securely.
Sometimes the answer will be AI-assisted human service. A customer dealing with a billing dispute, a delayed shipment, a healthcare question or a financial hardship may need an associate who can listen, interpret nuance and make a judgment call. In those moments, AI can still play a valuable role by summarizing history, recommending next best actions, identifying compliance requirements or reducing after-call work. But the human being remains central to trust.
And sometimes the answer will be proactive intervention. If data shows that a customer is likely to call because of a known issue, the best experience may be to notify them before they reach out at all.
That is what it really means to apply AI well. Don’t just automate; design experiences around customer intent, business value and the level of trust required in the moment.
I believe the future of customer experience is not isolated AI deployments, but orchestrated ecosystems where technology, operations and people continuously inform and improve one another.
AI is not the differentiator. The differentiator is how well companies connect technology, data and human expertise into one system that benefits the customer.
