Your Customer Is the Strategy, Not AI

Originally published on LinkedIn on May 7, 2026.

AI is everywhere right now, but in many conversations something else is starting to show up: fatigue.

That became clear to me even two years ago at the Customer Contact Week conference in Las Vegas. Like most industry events right now, nearly every session was centered on AI. When our session started, I asked a simple question: “Is anyone else tired of being sold AI right now?” The room laughed.

That moment changed the direction of the discussion. People started talking about what happens after it goes live. Things like how it fits into the operation, how you know if it’s working, and what you do when it doesn’t behave the way you expected. Even the conversation shifted to how you do QA and coach a bot.

What was clear to me was that no one needed another platform. They needed a way to make what they have work better.

Right now, the focus remains heavily on deployment, while the harder work of operationalizing AI is still underdeveloped. That gap is not new. Max Armbruster made the point recently that the BPO industry has been declared “dead” multiple times over the past four decades. Each wave of technology was expected to replace the model. Offshore delivery in the 1990s, political pressure in the 2000s, automation in the 2010s, and now AI. The tools change. The expectations don’t. Clients still need someone to take ownership of outcomes.

In many cases, AI is layered onto existing processes rather than used to rethink them. Agent summarization is a good example. It can save a few seconds at the end of a call. That has value and can even add up to substantive cost savings. But the customer still had to call, and in the end, the cost to serve didn’t really change.

This is the difference between optimizing a step and redesigning the system.

James Goldberg, SVP of Business Development and Client Success at Percepta, shared a situation that illustrates this clearly. While traveling, he realized he had forgotten essential medication and needed a prescription transferred quickly. What should have been a simple request turned into a multi-day process with multiple calls and conflicting system messages. Two organizations operating on different systems that could not coordinate with each other. By the time it was resolved, it required significantly more effort from the customer, the pharmacy, and the provider than if a single interaction had been handled correctly at the start.

What happened to Jamie was not just a system failure. It is what happens when a one-size-fits-all solution is applied without enough consideration for who the customer is and what they need in that moment. When the process broke down, there was no clear path to fix it quickly. What was intended to streamline the experience ultimately created more friction, more cost, and more work across the system. The stakes were clear: a customer needed medication, and the process required speed, coordination, and certainty. Instead, it delivered friction and delay.

Focusing on AI purely as a cost solution, rather than understanding it as part of the customer journey, puts loyal customer bases at risk. In automotive, the customer base itself should shape how AI is applied. According to S&P Global Mobility, adults aged 55 and over now account for nearly half of new vehicle registrations. These are customers making significant purchases and managing increasingly complex service needs. In many cases, they are not looking to be deflected. They are looking for clarity, confidence, and resolution.

That preference shows up in behavior. Research from CDK Global indicates that 66% of Baby Boomers prefer to call when scheduling service, compared to 31% of Gen Z. More broadly, a consumer study from InBusiness PHX found that 86% of customers prefer interacting with a human when they have a problem. When high-value customers expect a human interaction, the risk is not just deflection, it is misalignment.

Much of the current AI conversation assumes a level of readiness that is not always there. The focus shifts quickly to what can be automated rather than whether it actually improves the experience. In some cases, self-service works well. In others, especially when the situation is complex or high-stress, it creates more friction instead of less. A better approach starts with the customer and works backward, understanding who is calling, why they are calling, and what outcome they are trying to achieve, then applying a system of AI tools that work together to simplify the experience rather than complicate it.

At Percepta , much of our customer journey work has been grounded in that principle. Between 2019 and 2021, we partnered with a global automotive client to redesign their case handling model around a simple idea: clear ownership from start to finish. Customers were being passed across tiers, repeating information and waiting for updates, which created a fragmented experience. Rather than optimize the handoffs, we removed them. One associate owned the case end to end, supported by training, quality, and workflows aligned to that model. The impact was sustained, including a 60-point improvement in global NPS over three years, and the same approach has since been applied in markets like Germany in 2024 and 2025, which experienced an immediate 27-point NPS boost.

This redesign had nothing to do with AI. That is the point. It reflects the discipline required to apply AI well. Start with the customer, design the system around them, and then determine where technology adds value.

AI will continue to evolve. That much is clear. What will differentiate organizations is how deliberately they choose to use it.