Why Most Companies Have Fallen Behind on Customer Experience

For most of the last two decades, the conversation about customer experience (CX) in the boardroom followed a predictable arc. Efficiency was always the theme: Where are we on volume and headcount? What do the cost-per-contact numbers look like? And the ultimate question: How do we handle more with less?

That framing made sense for a world where CX was a cost center by design. The function existed to absorb customer problems and close them efficiently. Success meant fewer escalations, shorter handle times and a deflection rate that trended upward every quarter. The teams who ran these operations were good at it. They optimized hard for exactly what they were asked to optimize.

The problem is that the business changed underneath that model, and many companies have not updated their operating assumptions accordingly.

The shift has already happened

This is not a prediction. In most industries, the economics of the customer relationship have already moved in a direction that makes the old CX model structurally inadequate.

Customer acquisition has become significantly more expensive as digital advertising costs have risen and market saturation has increased across most categories. The customers companies already have are worth more. At the same time, the bar for abandoning one service for another has dropped. Competitors are easier to find, trials are lower friction, and a customer who leaves rarely comes back.

What that means in practice is that every customer interaction carries more financial weight than it did 10 years ago. A high-value customer who contacts support and has a bad experience is not a ticket. They are a churn risk with a name and a lifetime value attached to it. The organization that treats that interaction like a cost to be minimized is playing a different game than the one that treats it like a retention moment.

Most CX operations are still playing the old game. Not because the leaders running them do not understand the economics, but because the systems, the metrics and the org structures around them were built for a different time and a different purpose.

What AI actually changed

The case for CX as a revenue function is not new. Plenty of smart business leaders have recognized the connection between CX and business growth over the years. What is new is that the operational infrastructure to actually run CX that way now exists. 

AI is driving that shift, but not in the way that the loudest conversations about automation truly acknowledge.

It all comes back to that stubborn framing of cost efficiency above all. So much attention has been paid to how CX teams can deflect more contacts with chatbots. A valuable use case, no doubt, but still an outdated mode of thinking about CX’s potential. The real change AI made possible is closing a loop that was never closable at scale before.

Here’s what I mean: Imagine a customer who has been with you for four years contacts support about a billing discrepancy. In the old model, that ticket entered a queue, the billing issue got resolved, and the ticket got closed. Meanwhile, your retention team puzzled over why churn among long-tenure customers ran high last quarter. Nobody connected the two things, because the system was not built to connect them.

In the new model, that same contact surfaces more than just a billing issue. Your system recognizes a customer who has a high lifetime value, has contacted support twice in the past four weeks, and is showing signals associated with churn risk. The conversation routes to someone equipped to handle it, and that agent enters the conversation with the customer’s full context in front of them. They can make an informed decision on how to give the customer what they actually need and course-correct a churn risk in real time. And after the interaction closes, the system continuously analyzes its impact.

That is the loop: real-time risk recognition, intelligent routing, context in hand before the conversation starts, outcomes tracked on the back end. AI has made each of these steps operationally achievable. And the companies that view CX as a revenue function have gotten a head start.

What the new model looks like

Running CX this way requires more than the right technology. The function itself has to be built differently. Three things tend to separate the organizations that are making this work from the ones that are not.

The data layer is the foundation: The scenario I described above only works if AI and human agents already know the customer before the conversation starts — not piecing together context from three different systems mid-call, not asking the customer to repeat history they have already shared. That requires a unified data layer as the core foundation of CX. There’s not an AI tool that exists that can meaningfully transform a CX operation that’s built on fragmented, inconsistent data.

The team structure has to reflect the job: Not every interaction carries the same stakes. Some are routine and appropriately handled at volume. Others are moments where the right decision can protect a significant customer relationship. The teams that perform well on retention tend to put people in place who are specifically accountable for those high-stakes interactions and measured on outcomes, not throughput. That is a different role than a traditional support agent, and it needs to be staffed and managed accordingly.

Where CX sits in the org chart signals what it is expected to do: A function that reports into operations is being asked to manage cost and efficiency. One that reports into revenue leadership is being asked to manage customer outcomes. That reporting line shapes everything: the questions the function gets asked, the budget conversations it joins, the metrics it is held to over time.

A platform built on unified data is the core of CX transformation, the foundational layer that everything else is built on. But the people and the processes are just as critical. 

The metrics that belong in a CFO conversation

One of the consistent frustrations I hear from CX leaders is that they cannot get budget for the investments they know would make a difference. They know the scenario I explained above is possible, but they also know that their system is fragmented and disconnected data is the number one barrier to genuine CX transformation. But when they go to finance, the conversation stalls. 

The reason it stalls is usually a metrics problem. CX leaders often walk into that meeting with the wrong numbers, and so yet again, there’s nothing concrete to talk about except for cost efficiency.

Handle time and deflection rates speak a language finance recognizes, but they make a cost argument, not a revenue argument. If CX wants to be funded as a revenue function, it needs to make a revenue case. That means different data: revenue protected through CX intervention, churn attributed to unresolved support interactions, lifetime value correlated to resolution quality.

This might seem like a chicken/egg problem: Producing these numbers requires connected data infrastructure, and connected data infrastructure might be exactly what you are trying to get budget to invest in. But you don’t need your ideal measurement system before you walk into that meeting; what you need is enough directional signal to make the patterns visible.

Pull 12 months of churned accounts and look at their support history. Run the correlation on one product line. Take one quarter of resolution data and match it against renewal rates in the following year. The numbers will be rough, and that’s fine. It’s all about surfacing the patterns in language the CFO understands.

The framing that works is not “fund CX, and we can actually drive revenue.” It is: “Here is what I can see with the fragmented data we have today, and here is what I would be able to measure, track and act on with the right system.”

That reframes the budget ask from a cost center requesting more resources to a revenue function asking for the tools to prove its own value. Those are very different conversations.

The cost of waiting

The businesses that are seeing the greatest ROI from CX tend to share one characteristic: They made a decision, somewhere in the last few years, that CX would be accountable for customer outcomes rather than operational activity. Then they implemented the infrastructure to support that accountability, including the data layer, the team model and the metrics.

That decision may seem like a nice vision that’s too much of an overhaul to realize today. But businesses that have already shifted their approach to CX are pulling away from their competitors. For CX orgs who still work with yesterday’s reactive cost center framing, every quarter that passes is one where lasting customer relationships are being sacrificed for outdated operational metrics.

For most of the last two decades, the conversation about customer experience (CX) in the boardroom followed a predictable arc. Efficiency was always the theme: Where are we on volume and headcount? What do the cost-per-contact numbers look like? And the ultimate question: How do we handle more with less?

That framing made sense for a world where CX was a cost center by design. The function existed to absorb customer problems and close them efficiently. Success meant fewer escalations, shorter handle times and a deflection rate that trended upward every quarter. The teams who ran these operations were good at it. They optimized hard for exactly what they were asked to optimize.

The problem is that the business changed underneath that model, and many companies have not updated their operating assumptions accordingly.