Why Not Every Guest Complaint Is Equally Fixable

Why doesn't a low sentiment score tell you what to do?

Take three themes from the same anonymized property.

Location & Access is 79% positive. Housekeeping & Cleanliness is 42% positive. Bathroom Facilities is 52% positive.

A GM looking at those numbers would reasonably rank Housekeeping first, then Bathrooms, then Location. Lowest score, biggest problem. But the three aren't the same kind of problem.

Housekeeping sits in Service & Staff Experience, a pillar the dashboard describes as flexible and operational, funded from the operating budget. Bathroom Facilities sits in Physical Product & Asset Condition, which is capex-weighted and belongs in a capital plan. Location & Access is structural. Management can't move the building.

Same kind of data, three different decisions, and three different people in the room. A sentiment percentage can't tell you that. The type of problem can.

The previous article in this series showed where the revenue opportunity sits - Which Part of Your Hotel Is Holding Back the Most Revenue? This one asks a different question: what kind of problem is each part of it, and who decides?

Where do the numbers come from?

A quick recap, so this article stands on its own.

The example is an anonymized 156-room property in the Lake Geneva region. The themes, mention counts and positive-sentiment percentages are real and review-derived, taken from the property's public guest reviews. The operating inputs are illustrative: 156 rooms, an ADR of CHF 245.00, 68.5% occupancy and a RevPAR of CHF 167.82. Public reviews don't contain these, and the hotel didn't supply them.

Each theme's gap to a healthy 85% benchmark is weighted by its share of mentions to give a theme score lift. Theme lifts add up to pillar scores, and pillar scores add up to a hotel-level Revenue Opportunity Score (ROS) of 8.76 points. Published Cornell hospitality research is then applied once, at hotel level, giving an illustrative annual opportunity of CHF 1.13M to CHF 1.25M. That total is split across the pillars in proportion to each pillar's share of the lift.

Two things follow. There is no CHF figure for any single theme, so themes in this article are compared by score lift only. And the total is a ceiling scenario, not a forecast, because it assumes every gap closes to 85% at once. (New to the approach? What Is Guest Experience P&L Mapping? covers the basics, and From Sentiment Score to P&L walks through the calculation step by step.)

How do structural and operational hotel issues differ?

For every pillar, the dashboard asks three questions.

Controllability: can management change this, and how? The answer might be flexible, partly structural or structural.

Decision: whose call is it, and which budget does it come from?

Implication: what does acting on it actually affect: asset value, guest retention, positioning?

Here is how the five groupings in the example answer them.

The pillar ranges are slices of one hotel-level estimate. They add up to the headline range by construction. They aren't four independent forecasts.

What the table really shows is that “improve guest sentiment” isn't one task. It's at least four, owned by different people, funded differently, and judged on different timescales.

The structural-versus-operational split matters most for the people reading the number. A GM can act on the operating-budget pillars within a single budget cycle. An owner or investor decides the capital-weighted one. And for the structural one, nobody at the property can act on the underlying cause at all.

Hotel location sentiment: a score you read, not fix

Location & Access is the most-mentioned theme in the entire dataset, with 1,344 mentions. It's 79% positive, a 6-point gap to the healthy benchmark.

If it were counted, it would carry the largest weight in the model and the second-highest theme score lift, at 1.42, behind only Maintenance & Upkeep (1.70). Yet it contributes nothing to the CHF 1.13M–1.25M figure.

That's deliberate. The dashboard classifies it as structural: not actionable by management. Treating it as revenue a GM could capture would overstate what the property can do.

But leaving it out of the number isn't the same as ignoring it. Hotel location sentiment still carries information, and the dashboard says what kind. The decision attached to it is repositioning and expectation-setting: which guest segment the property targets, how guests' expectations are set before they arrive, and how pricing sits relative to the location. The implication is a long-term asset-value and strategy question, relevant to owners and investors rather than to the people running the property day to day.

That's a useful distinction in practice. A 79% location score read by a GM as “something to fix” produces frustration. Read by an owner as “something to price and position around,” it produces a decision.

The wording here is careful. The dashboard doesn't say part of Location is fixable. It says the theme is structural, and it points to a different kind of response. This article follows the same line.

What are the two warnings built into the model?

Two other classifications are easy to miss, and each prevents a specific mistake.

Value for Money is often a symptom, not a root cause. In the example it sits at 50% positive across 193 mentions, a 35-point gap and a score lift of 1.19. It would be tempting to treat that as its own project. The dashboard says otherwise: guests who rate value poorly are often really rating the product against the price they paid. Before treating it as a separate fix, check whether it moves once the Physical Product gaps close. The pillar is classed as partly structural, and the decision is pricing and positioning only after that check.

Food & Beverage shouldn't be over-weighted just because it's mentioned often.

Breakfast Quality has 513 mentions and sits at 82% positive. Food & Beverage has 291 mentions and sits at 80%. Together they add just 0.53 points of lift, about 6% of the total. The dashboard calls this a near-term, minor adjustment and names it explicitly so it doesn't get over-weighted simply because guests talk about it.

Both warnings come from the same principle. Volume and score can mislead on their own. What matters is what kind of problem a theme represents.

Hotel capex prioritization: how do you sequence the biggest pillar?

Physical Product & Asset Condition holds roughly three-fifths of the opportunity, so it's where hotel capex prioritization becomes a real question. Capital is finite. Not everything in the pillar can be funded at once, and order matters.

The pillar has five themes.

Three of the five do most of the work. Maintenance & Upkeep, Room Quality and Bathroom Facilities contribute 4.46 of the pillar's 5.33 points. That already narrows the conversation. Noise Levels and Bed Comfort matter, but they carry far less of the pillar's weight.

The themes also have different shapes. Room Quality is a large conversation with a modest gap: 868 mentions, 9 points below healthy. Bathroom Facilities is a smaller conversation with a deep gap: 237 mentions, 33 points below. Both give the same score lift of 1.38, yet a broad, shallow problem and a narrow, deep one rarely call for the same scope of work.

The dashboard also ties this pillar to something beyond the guest score. It directly affects asset value and renovation ROI, which makes it the pillar most relevant to an owner or investor evaluating the property, not just the GM running it.

One caution matters here. A theme inside a capex-weighted pillar isn't necessarily a capital project. Maintenance & Upkeep complaints might include routine repair and upkeep work as well as items that need real investment. Bathroom Facilities might include both a refurbishment question and day-to-day condition. Splitting each theme into “routine” and “capital” is a judgment for the people who know the property. It's not a dashboard output, and SENTEEZ holds no intervention-cost data.

So the model does something narrower than telling you what to build first. It shows which themes carry the most weight inside the pillar, tells you the pillar is a capital-plan decision, and leaves the scoping to the people who can see the building.

How should you read this honestly?

The same boundaries that apply to the rest of the series apply here, plus one more.

It's a ceiling. The total assumes every gap closes to 85% at the same time. Applying research relationships to a move that large is a stretch, and no hotel should plan around the headline figure.

Pillars overlap. Guest feedback doesn't fall into tidy boxes. Maintenance, Bathroom Facilities and Room Quality often describe related problems, so the slices aren't perfectly clean.

Opportunity isn't attribution. The score shows where to look first. It doesn't prove what a specific fix will deliver.

The range is a presentation band. The spread around the midpoint is a display choice, not a statistical confidence interval.

Controllability labels classify who decides, not how long or how costly a fix is. “Operational” doesn't mean cheap or quick, and “capex-weighted” doesn't mean every item needs capital. The labels point to the right decision-maker. They don't price the decision.

Before your next capital meeting

You don't need a model to start. For each theme or pillar in your own review data, ask three questions.

1. Is it structural or flexible? Can anyone at the property change the underlying cause, or is it a fixed feature to price and position around?
2. Which budget owns it? Capital plan, operating budget, or pricing and positioning?
3. Is it a symptom of something upstream? Poor value ratings, for instance, may be pointing at the product rather than the price.

Then sort your list by two things together: how much opportunity sits behind each item, and who owns the decision. The result is rarely the same as ranking by sentiment score. That gap is where a more useful conversation begins, and it often means different people need to be in the same meeting.

Where does this go next?

Industry peers have pointed out a fair limit of any opportunity score: it shows where to look, but not what happened after you acted. We're exploring how to track results after a fix, comparing real outcomes against the pre-intervention estimate. It's early thinking rather than a finished feature, but it's a natural next layer on top of the opportunity signal.

The bottom line

The biggest number isn't always the first move, and some numbers aren't a move at all.

In this anonymized example, one pillar is a capital-plan decision, one is an operating-budget decision, one is a pricing check that starts with the product, one is a minor adjustment, and one is a structural factor that management can't act on, so it's shown separately. A sentiment score alone would have treated them all as the same kind of problem.

The Revenue Opportunity Score shows where to look first. Controllability shows who acts, and with which budget.