Your Reports Are Beautiful and Logically Sound, But They Don't Tell You Why Guests Stopped Coming

In the previous article, we made one point: revenue is a lagging indicator. Every report in your hands describes the past.

So you decide to start looking for “reasons.” You open your toolbox. Inside, there are three things.

Method One: Read Financial Reports

This is the most intuitive. Income statements, cash flow statements, cost structure analysis. You see F&B revenue down 12%, labor costs up 8%. Conclusion: there’s a problem with F&B efficiency; we need to control headcount.

The advantages of financial reports are obvious: systematic, comparable, historically tracked. Public company CFOs rely on them for decisions. Investors rely on them to judge value.

But financial reports have a structural blind spot: they are a summary of outcomes, not a record of causes.

The financial report tells you “F&B revenue declined 12%.” It doesn’t tell you: which tables of guests didn’t show up? Was it because the food was cold last time? Was it because the parking spaces were removed? Was it because a new place opened next door?

There’s a line item on the income statement called “Operating Revenue.” That number is the sum of thousands of independent consumption decisions. When the number drops, you see the aggregate, but you can’t see the reason behind any single decision.

Using outcomes to search for causes is like using an autopsy report to do an epidemiological investigation. The cause of death is written clearly, but the transmission pathway? Unknown.

Method Two: Track Orders

This is more real-time than financial reports. You check the reservation system daily: how many rooms sold today, what’s the ADR, how does Pace compare to last year.

The advantage of order tracking: real-time, specific, traceable to every transaction.

But order data has a more insidious blind spot: survivorship bias.

Your Order Data Is Those Returning Bombers

Every order comes from a guest who “decided to choose you.” Those who “considered you but chose a competitor” leave no order record. They’re not in your system. The cleaner your data, the worse the survivorship bias — because you can only see those who stayed.

The problem with order tracking isn’t insufficient data. It’s that you’re always analyzing a group of people who already chose you, then using their behavior to infer why non-visitors don’t come. Logically, this doesn’t hold up.

Let’s make this more concrete. Imagine you have 1,000 potential guests who considered your hotel in the past three months. Of those, 300 booked a room and appear in your reservation system. 700 didn’t book. They vanished without a trace.

Your reports analyze those 300 people. You analyze their check-in times, spending amounts, return rates, and complaint records. You get a set of elegant charts and insights.

But what about those 700 people? Where did they go? Who did they choose? Why didn’t they choose you?

They’re not on your reports. They’re not in your system. They are a group of “invisible rejecters” — they made a choice, but that choice (not choosing you) leaves no record in any system.

Wald’s insight holds true today: you reinforce where you have data (the wings), but you need to reinforce where you don’t have data (the engines). In the hotel industry, the engines are those 700 non-visitors. And how do you see the bullet holes on the engines? You can’t go ask 700 strangers “why didn’t you choose me.”

But they’re already telling you. Just not through your system.

Method Three: Conduct Market Research

OK, you say, then I’ll just ask guests directly.

NPS surveys, CSAT satisfaction surveys, focus groups. Send out 500 questionnaires, get 120 back, analyze the results, get a set of scores and recommendations.

The advantage of market research: direct, structured, can quantify “satisfaction.”

But market research has three fatal structural problems:

First, response bias. People willing to fill out surveys are not a random sample to begin with. Satisfied people have an incentive to fill them out (to express gratitude), extremely dissatisfied people also have an incentive (to vent). But the “thought it was fine, but won’t come back next time” group — the largest group — won’t fill them out. Your survey results systematically ignore the middle ground.

Second, social desirability bias. When a survey asks “would you recommend this hotel?” respondents subconsciously skew toward the positive. Not because they’re lying, but because humans automatically adjust their answers in formal survey settings.

Third, frequency limitations. Conducting market research once a quarter is already considered frequent. But guest sentiment changes every week. Using quarterly snapshots for daily decisions gives you far too little resolution.

Frederick Reichheld — the inventor of NPS — when he published the concept in Harvard Business Review in 2003, claimed it could “predict growth with a single number.” A decade later, Reichheld himself admitted: NPS’s predictive power in the service industry is far lower than in retail. Why? Because service experiences are dynamic, subjective, and context-dependent — a 0-10 number can’t compress all of that.

The Common Blind Spot of All Three Methods

Let’s look at all three methods together:

Method	What It Sees	What It Misses
Financial reports	Summary of outcomes	The reason behind each transaction
Orders	Behavior of those who came	The voice of those who didn’t
Market research	Opinions of those willing to fill surveys	The true feelings of those who won’t

Three methods, three blind spots, but pointing to the same conclusion:

They’re all looking at “what happened.” None of them is looking at “why it happened.”

More precisely: they’re all looking at after-the-fact numbers and after-the-fact responses. None of these methods captured the guest’s true feelings at the moment they made their decision.

Your Guests Have Been Telling You All Along

There is one information source that simultaneously possesses three qualities: real-time, unvarnished, and voluntary.

Not surveys (you asked for those; they’re not necessarily voluntary). Not orders (those are outcomes, not feelings). Not financial reports (those are summaries, not originals).

It’s your guests, after leaving your establishment, voluntarily, publicly, in their own words, telling the world what they experienced.

This information is generated every day. It’s real-time — guests write it the same day or the day after their experience. It’s unvarnished — no one reviews it, no survey format constrains it. It’s voluntary — no one asked them to write it; they wrote it because they felt it was worth saying.

The question is: are you reading it? Do you have a systematic method to turn these scattered, seemingly subjective, emotion-laden voices into actionable decision signals?

Or are you like that general manager — seeing April’s reports in June, with all three methods in hand pointing to the past, while the guests’ voices have been there all along. He just wasn’t listening.

Where are these voices? What do they look like? How do you read three months of future revenue trends from them?

That’s the next question to answer.

---

Download the Blind Spot Checklist + Leading Indicators Starter

Leave your email to unlock two pages of practical resources: page one is a blind spot comparison table for all three methods (with self-assessment questions), and page two tells you three leading indicator signals you can start observing right away.

The moment you get it, a thought might flash through your mind: “If I have to check all this manually every week, who has the time?”

That thought is the starting point of the problem.

---

Further Reading

← Previous: Why Is It Already Too Late by the Time You See Revenue Decline?

Revenue is a lagging indicator. Every report in your hands describes the past. This article breaks down why all standard KPIs are retrospective.

→ Read the previous article