This article was based on the interview with Alyse Fuller, Customer Experience Program Manager at United Rentals by Greg Kihlström, Marketing AI adoption keynote speaker for The Agile Brand with Greg Kihlström podcast. Listen to the original episode here:
For years, the Net Promoter Score (NPS) has been the North Star for customer experience programs. We’ve built dashboards, tied bonuses to it, and presented its steady upward climb in boardrooms as definitive proof of our success. And for good reason—it’s a simple, powerful metric. But as any seasoned leader knows, the comfort of a high, stable score can be deceptive. It can become a vanity metric, a signal so consistent that it lulls us into a state of operational complacency, masking the subtle friction points and emerging issues that lie just beneath the surface. When that score finally does shift, the resulting scramble to understand why often reveals that we’ve been measuring the outcome, not the drivers.
This is the central challenge facing modern marketing and CX leaders: moving our organizations beyond the score. It’s about shifting the discipline from a reactive, number-chasing exercise to a proactive, problem-solving engine. The good news is that we now have a uniquely powerful ally in this effort. Generative AI is not just another tool for analysis; it’s a force multiplier that can close the gap between insight and action at an unprecedented scale. By synthesizing vast and disparate datasets—from customer verbatims to frontline employee notes—AI allows us to move from symptom identification to root cause analysis, transforming our roles from data gatekeepers to strategic advisors. The experience of United Rentals, a massive organization with a remarkably lean CX team, offers a practical blueprint for how to make this transition a reality.
The Danger of a High-Performing Program
The paradox of a successful CX program is that prolonged success can make it difficult to innovate. When scores are consistently high, as United Rentals’ were for years, the organizational appetite for deep, systemic analysis wanes. The “if it ain’t broke, don’t fix it” mentality prevails. The problem, as Elise Fuller discovered, is that this stability can starve your insights engine of the very thing it needs to get stronger: variance. Without new signals or unexpected changes, you can’t effectively pressure-test your assumptions or uncover latent issues that may be slowly eroding the customer relationship. A sudden dip in a long-stable score, while alarming, is also an opportunity.
“There had been such consistency there, it makes it difficult to really pressure test your insights engine. You need some variety and you need some new signals and things to really see, will all these rules and all these conditions hold up no matter what the circumstances are? Or is it possible there’s something lurking under the surface that I haven’t been measuring all this time?”
Fuller’s point gets to the heart of a critical leadership responsibility: cultivating a healthy skepticism of our own data. A consistently high NPS can mean customers are truly delighted, or it can mean we aren’t asking the right questions of the right people. It might reflect a great product but mask a cumbersome delivery process that customers have simply learned to tolerate. When the catalyzing moment of a score change arrived for United Rentals, it wasn’t just a problem to be solved; it was a mandate to look deeper. It forced a transition from a 2-Why analysis to a 5-Why analysis, pushing past the customer-facing symptoms to uncover the operational realities behind them. This is where the real work begins, and where a purely customer-centric dataset falls short.
Connecting the Frontline to the Bottom Line
Customers are experts on their own experience, but they are not experts on your business operations. They can tell you a delivery was late (the symptom), but they can’t tell you it was because of an inefficient dispatch system or a breakdown in communication between sales and logistics (the root cause). To truly understand the ‘why,’ we must bridge the gap between the external voice of the customer and the internal voice of the employee—the people “behind the curtain” who live the operational complexities every day. For United Rentals, this meant synthesizing customer feedback with notes from frontline team huddles, creating a holistic view of the customer journey from both sides of the rental counter.
“The customers can talk very well about the symptoms that they’re experiencing… they don’t know all of the logistics. So we really had to look at or get input from people who are already behind the curtain and they know all the details of all of our systems and what it really takes to deliver that kind of experience.”
This is where AI demonstrates its true power. Manually correlating thousands of customer comments with thousands of internal employee notes is an impossible task for a lean team. But for a large language model, it’s simply a matter of processing structured and unstructured data. By feeding both datasets into its platform, United Rentals could see, in plain language, how internal challenges directly impacted customer sentiment. Phrases like “customers want us to focus on these things” were directly paired with “our teams are talking about these things.” This connection is transformative. It turns CX from a marketing-owned metric into a cross-functional operational imperative, giving leaders from logistics, sales, and service a shared language and a clear line of sight into how their decisions affect the end customer.
Adoption Through Utility, Not Hype
The temptation with any new technology, especially one as hyped as AI, is to launch it with fanfare. We schedule company-wide trainings, create slick announcement decks, and celebrate the novelty of our shiny new tool. But this approach often backfires. It can create anxiety among frontline users who fear being replaced or burdened with a complex new process. Fuller and her team at United Rentals took a refreshingly counterintuitive approach when rolling out Medallia’s AI-powered Smart Response tool: they didn’t. There was no formal announcement, no mandatory training—they just turned it on.
“A big resounding theme is it just has to work… It’s supposed to be integrated into what we’re already doing anyway and make things easier and more efficient. So when thinking about smart response… it’s an easy button. It should not require a big announcement or too much fanfare, which… can be off-putting to some people when really they could benefit from the tool. So let’s just make the tool available and see what happens from there.”
This “utility over novelty” strategy is a masterclass in change management. By embedding an AI feature directly into an existing, familiar workflow, it becomes a natural extension of the process rather than a disruptive new mandate. The tool’s value had to speak for itself. If it genuinely made a manager’s job easier by providing a better, more personalized starting point for a customer response, they would use it. If it didn’t, they wouldn’t. The result was a steady, organic adoption curve driven by proven value, not by corporate hype. This approach minimizes friction and builds trust, demonstrating that AI is there to augment their capabilities, not to add complexity or threaten their role. It’s a powerful lesson for any leader tasked with driving digital transformation: the most effective tools are the ones that feel invisible.
From Analyst to Strategic Partner
The ultimate promise of AI in customer experience is that it shortens the distance between listening and acting. For United Rentals, the impact was immediate and measurable: frontline managers were closing customer alerts six hours faster. But the more profound impact was on the role of the CX leader itself. When the platform automates the time-consuming work of data aggregation, correlation, and summarization—the “proving of the problem”—it liberates the human expert to focus on higher-value work. Time once spent buried in spreadsheets and manually building reports is repurposed for strategic planning and cross-functional consultation.
“That time spent on the analyst side is what’s been greatly reduced. I can generate these reports… and then spend my time more on discussing what we’re seeing here, gathering other context… bringing more just a more robust nature to what that conversation can sound like.”
This is the end game. AI doesn’t make the CX leader obsolete; it elevates their function. By democratizing access to insights and presenting them in natural language, it empowers leaders across the organization to pull the information they need, when they need it, without having to wait for a quarterly report. The CX program manager is no longer the sole gatekeeper of insights but rather the facilitator of action. They become an internal consultant who can add context, challenge assumptions, and help operational teams design and implement meaningful improvements based on a shared, data-driven understanding of the customer’s reality.
The journey undertaken by United Rentals is a microcosm of the broader evolution happening across our field. We are moving from an era defined by measurement to one defined by action. The tools at our disposal, particularly generative AI, are fundamentally changing the calculus of what’s possible, even for the leanest of teams. By focusing AI on connecting disparate data sources and embedding its capabilities invisibly into existing workflows, we can empower our entire organization—from the C-suite to the frontline—to operate with a clearer, more immediate understanding of the customer.
This shift requires us, as leaders, to evolve as well. Our greatest value is no longer in our ability to analyze data, but in our ability to foster the conversations and drive the changes that the data demands. It’s about moving from being the chief analyst to the chief enabler, arming our colleagues with the insights they need to make smarter decisions faster. The goal isn’t just to achieve a higher score; it’s to build a more responsive, customer-centric organization from the inside out. And that, in the end, is a far more meaningful measure of success.
