AI isn’t the answer to every question, but it can be the key to unlocking operational transformation

“What is, AI?” could be the answer to every Jeopardy question being thrown around boardroom tables today.   

But turning AI from a headline into actual outcomes—especially during times of operational transformation—requires more than just good ideas. 

Leaders often get frustrated when a cutting-edge AI tool delivers answers that don’t match their trusted spreadsheets or years of on-the-ground experience. To deliver on AI’s promise, we need to pair these new capabilities with clear metrics, best practices, and real accountability. That starts with one simple question: What exactly are we asking these tools to do? 

The “unique unicorn” problem in operational transformation

Through my years working across a global hospitality business, I’ve come to understand one truth: every hotel thinks it’s a unique unicorn. Try to benchmark guest experience across a portfolio of properties, and operators push back—“We’re different.”

In many ways, they’re right. One might be tucked next to a stadium or a highway, another inside a centuries-old building with an equally historic HVAC system. Others have unique hotel teams that may face local labor laws, union dynamics, or language barriers. And to be fair, offering a unique, “one of a kind” experience in the hospitality industry is a goal worth chasing.

So it can be a challenge in large, franchised Fortune 500 businesses to benchmark across properties in a quest to drive improvement. 

During one recent critical transformation, when faced with this problem, we flipped the script. We set a universal goal of holding hotels accountable to improving guest experience, measured by NPS, while respecting that every property might need their unique path to get there. 

That’s where AI came in.

We used AI to mine a treasure trove of qualitative and quantitative guest feedback to guide operators on their improvement journey, developing a framework for improvement, leveraging AI.

What IIF? The Intelligent Improvement Framework (IIF): A scalable, human-centered approach

To guide our operators, we built an “Intelligent Improvement Framework.” It worked, and fast. Within six months, we saw statistically significant improvements across a network of largely franchised properties.

The “What IIF” approach has four critical parts: 

1. Pre-Work: Use AI to Turn Feedback into Focus. This is where the “Intelligence” part of AI makes life faster and easier. Tens of thousands (or millions) of guest comments cannot be easily parsed by a busy hotel operator, who is making beds, fixing meals and checking guests in. This foundational step is building an algorithm or AI model that can reliably, mathematically, associate thousands of collected data points with a single net promoter score, and analyze and identify the top three factors most likely to improve that score. 
2. Serve it Up, Simply: Each property received its top three focus areas for improvement. No digging required. Operators immediately recognized familiar themes from their guests, but now with confidence and clarity—and without the burden of analysis. These focus areas, customized to their “unique unicorn” of a property, now help them improve their guest experiences, based on intelligent analysis of past scores.
3. Connect to Actionable Tools: Each of the 3 focus areas is linked through to practical resources in a Knowledge Library: checklists, how-to guides, and best practices, all intended to drive to a specific priority. With a light-touch, “root cause analysis” activities are layered in, targeted to the skill level of the user.  
4. Build an Action Planning Portal: With their roadmap in hand of the top three areas to drive improvement, operators could assign tasks, allocate resources, and set deadlines within the Action Planning Portal. For example, to improve on the guest pool experience, a resort manager might be tasked with executing the corporate “Pool Checklist” by month’s end. Visibility and transparency at the regional and corporate level ensure accountability—and when NPS goes up (or doesn’t), the team can quickly course-correct. It’s a framework, not a formula – it only works when the team is accountable. 

From hospitality to QSR, retail, manufacturing, and healthcare: Scaling the “What IFF” approach

This same model can translate to other franchise environments where customer experience is a driver of success—quick service restaurants, retail chains, or any B2C enterprise with localized operations and customer feedback. It can also be applied in situations where customer experience is not the key factor but where many different components need to be analyzed and optimized separately.

• QSR: A franchise sees NPS drop. AI flags “Renovation” as a key issue. Using a checklist from the Knowledge Library, the team identifies needed upgrades and builds an ROI business case for capital investment, showing the statistical relationship between declining customer scores and declining revenue.
  
• Retail: Weekend shoppers consistently leave poor feedback scores on a touch pad as they leave a retail store during weekend afternoons. AI analysis from follow-up surveys and operational data inputs (inventory, staffing) reveals that store cleanliness is one key factor driving the guest NPS. A weekend manager gets a simple checklist, assigned via the platform, with accountability and visibility across the management team.  
• Call Center: A business relies on its call center to triage customer calls. While they already have intelligent call routing, the team is seeing low satisfaction on “high urgency” calls, and provides call responders with tools to analyze root causes, and improve both speed and customer satisfaction at the same time.
• Healthcare: A healthcare business uses skilled nurses to analyze complex hospital bills, hundreds of pages long. Trained AI can be used to “read” those bills, and offer up suggestions of where those bills may need to be more closely analyzed. By tracking the rate of success on the suggestions, and improving the feedback loop for nurse reviewers, the model can be refined to improve accuracy, time, and value – and these statistics can be visible and accountable across the leadership team. 
• Manufacturing: A manufacturing plant has 150 different machines making components of DME (durable medical equipment). Different machines experience different levels of downtime, error rates and overall quality. Team builds an AI model that analyzes the underlying data and desired KPIs, by machine; each machine can then be optimized for its desired quality KPIs – and robust tracking and transparency gives leaders a daily dashboard to monitor. 

How “unique unicorns” win

The “What IIF” approach works because it respects local nuance while scaling best practices. Operators feel heard, supported, and empowered, not measured by a one-size-fits-all yardstick. That balance creates trust across the enterprise, from front-line staff to regional leaders to the C-suite.

AI can be transformative—but only when paired with strong execution and a deep understanding of context. When you combine creativity with accountability, even the most “unique” operations can improve, at scale.