Win, Lose, or Doesn’t Matter: How to Read a One-Competitor Hotel Benchmark

Key findings at a glance

1. Two Lucerne hotels rated 4.46 and 4.52 have overall weighted sentiment of 87.2% and 88.4%, a 1.2-point difference that hides aspect-level gaps ranging from +5 to −14 points.

2. The largest gap (Noise Levels, −14 points) rests on only 50 Review Mentions, about 3% of that hotel’s aspect conversation. A 4-point gap on Front Desk & Staff rests on 359 mentions and about 21.5% of it.

3. Across the 37 Lucerne hotels with at least 50 Review Mentions on Front Desk & Staff, the share of positive mentions runs from 44% to 95%.

4. Both hotels score about 51–52% on Value for Money, a tie, but Lucerne’s town figure is 44.4% across 29 hotels. A tie at a weak level is a shared market problem, not a win.

5. Every gap falls into one of three buckets: win (protect), lose (fix), or doesn’t matter yet (watch).

Two hotels, one rating band, very different stories

Imagine you run Hotel A. Your SENTEEZ rating is 4.46, built from 644 reviews. A competitor in the same town, Hotel B, sits at 4.52 from 664 reviews. Both are strong hotels in one of Switzerland’s busiest tourist cities, and the difference between them is six hundredths of a star. Most operators would call that a tie, or a rounding error, and move on.

That’s the scoreboard mistake. A rating squeezes everything guests said into one number, so two hotels can land in the same place for completely different reasons.

Lucerne shows how much a single number hides. Across the 39 hotels in our dataset, ratings run from 3.02 to 4.66. On Front Desk & Staff alone, the 37 hotels with at least 50 Review Mentions on that aspect range from 44% positive to 95% positive. A 51-point spread on one aspect simply can’t be read off a star rating.

This article is about the question a rating can’t answer: when you put your hotel next to one competitor, which differences deserve action? Not all of them do. Some are noise, some are decisive, and the biggest number on the page is often not the one that matters most.

What a one-competitor benchmark actually measures

A one-competitor benchmark takes your hotel and a single rival and lines them up aspect by aspect: Front Desk & Staff, Room Quality, Breakfast Quality, Value for Money and so on. The method behind it is called aspect-based sentiment analysis. In plain words, every guest comment gets sorted by topic, and each topic is scored on how many of its mentions are positive.

Three terms come up throughout:

• % Positive: the share of Review Mentions on an aspect that are positive.
• Review Mentions: how many times guests mentioned that aspect. These aren’t unique reviews, because one review can mention several aspects.
• Hotels Reporting: in a town benchmark, how many hotels have the aspect among their top eight topics. It measures how prominent a topic is, not whether it’s absent elsewhere.

The data comes only from publicly available reviews over a 12-month window, and the SENTEEZ rating is calculated from review content rather than copied from a booking platform. Every figure here describes what guests say, not what happens inside a hotel.

One practical point before the walkthrough: each hotel reports its top eight aspects. Hotel A and Hotel B happen to report the same eight, which is what makes this comparison clean. When an aspect shows up for one hotel but not the other, don’t compare it. “Not in the top eight” doesn’t mean “zero.”

The three buckets: win, lose, or doesn’t matter

Every aspect gap in a head-to-head falls into one of three buckets.

• Win (protect). You lead on an aspect that guests mention often. This is where your reputation lives. Protect it, and don’t trade it away to fund something else.
• Lose (fix). You trail on an aspect with enough evidence behind it and enough reach among your guests that it’s costing you. This goes to the top of the list.
• Doesn’t matter (yet). The gap is real on paper, but it rests on few Review Mentions or touches a small slice of your guest conversation. Watch it. Don’t build a quarter’s plan around it.

The word “yet” is deliberate. Low mention volume means “monitor,” not “ignore,” and a poor score on a thin aspect can become a thick one quickly. There’s also a fourth outcome, the tie. A tie at a high level is parity worth protecting. A tie at a low level is a shared weakness, and the walkthrough below has one.

The three-question test

To place any gap in a bucket, ask three questions in order.

Question 1: How big is the gap?

Start with the size of the difference in % Positive, but read it against the sample. As a working guideline, not a statistical rule: on a few hundred mentions, differences of two or three points are usually noise, and five or more start to look like signal. The smaller the sample, the bigger the gap needs to be before it means anything.

Question 2: How much evidence stands behind it?

Look at the Review Mentions behind each score. Aspects aren’t mentioned equally, even inside one town. In Lucerne, Location & Access draws 13,312 Review Mentions across 39 hotels, while Maintenance & Upkeep draws 1,640 across 22. At hotel level the same pattern holds. A score built on 450 mentions is a finding. A score built on 50 is an indication, and below that it’s a hint.

Question 3: How much of the guest conversation does it touch?

Finally, ask what share of this hotel’s guest conversation the aspect represents. Front Desk & Staff touches almost every stay. Noise Levels touches the guests whose nights it affects, which is a smaller group. For Hotel A, Staff is about 21.5% of all its aspect mentions and Noise is about 3%.

A quick check: multiply the gap by the aspect’s share of the hotel’s mentions. It’s a back-of-envelope estimate of how much each aspect contributes to the overall difference between the two hotels, and it keeps big numbers on thin aspects from stealing the headline.

Put it all together: a guest experience gap matters in proportion to its size, the number of Review Mentions behind it, and the share of the hotel’s guest conversation it touches.

Walkthrough: Hotel A vs. Hotel B

The table below is built from public review data for two real Lucerne hotels, anonymized as Hotel A and Hotel B. Neither is a SENTEEZ client. Hotel A (4.46, 644 reviews) is the hotel you’re reading the benchmark for. Hotel B (4.52, 664 reviews) is the rival. “Gap” is Hotel A minus Hotel B, and the town column shows the Lucerne benchmark across 39 hotels, with Hotels Reporting in brackets.

Illustrative anonymized example built from public review data. Review Mentions are not unique reviews. Overall weighted sentiment: Hotel A 87.2%, Hotel B 88.4%.

The loss to fix: Front Desk & Staff

The Staff gap is only four points, and on its own it’s borderline: on roughly 360 and 450 mentions, four points is suggestive rather than conclusive. What moves it into the “fix” bucket is questions two and three. Staff is the second most-mentioned aspect for Hotel A, about 21.5% of its conversation, and the gap isn’t alone. Room Quality and Breakfast Quality lean toward Hotel B by two points each. Together, those three aspects account for nearly half of Hotel A’s guest conversation, and all three point the same way. Consistency across the aspects guests talk about most is a stronger signal than any single gap.

By the back-of-envelope check, Staff alone accounts for roughly 0.9 points of the 1.2-point overall difference between the two hotels.

The win to protect: Food & Beverage

Hotel A leads by five points on Food & Beverage, with 214 mentions behind it and about 13% of its conversation. It’s a differentiator in a way Staff isn’t: only 20 of the 39 Lucerne hotels have F&B among their top eight topics, and the town figure is 81.6%. Hotel A’s 93% sits well above it. This is the kind of lead that gets cut by accident when budgets tighten.

The biggest gap that doesn’t matter yet: Noise Levels

Noise shows the widest gap on the page, 14 points, and it’s the number most likely to grab attention in a meeting. Look at what stands behind it. Hotel A’s score rests on exactly 50 Review Mentions, the same 50-mention floor SENTEEZ uses to decide whether an aspect is thick enough to benchmark at town level, and it represents about 3% of the hotel’s conversation. By the quick check it explains only about 0.4 points of the overall difference, less than half of what Staff explains.

There’s context in the town column too. Noise is a weak spot across Lucerne, at 51.0% across 30 hotels and 2,163 mentions. Hotel A’s 68% is 17 points above the market, so the “loss” to Hotel B is still a lead over the town. Put it on a watch list. It’s a reason to keep reading, not a reason to renovate. Bed Comfort (−5 points, 63 mentions, under 4% of the conversation) goes on the same list.

The tie that isn’t a win: Value for Money

Both hotels land at 51 to 52% on Value for Money, a one-point gap, so a scoreboard reading would call it a draw. But Lucerne’s town figure is 44.4% across 29 hotels, and 28 of those 29 hotels score below 70%. Neither hotel is winning here. They’re sharing a market-wide weakness. Whichever hotel moves first would gain a lead most of the town doesn’t have, which is a different kind of decision from “closing a gap with a competitor.”

See your own version of this table
The free SENTEEZ teaser report does this for your hotel and one competing hotel: your rating, a one-competitor benchmark showing where you win and lose, and your top recurring guest themes. It’s delivered within 48 hours, with no commitment.
Get your free teaser report →

Four common misreads

1. Benchmarking against the wrong rival

A benchmark is only as useful as the hotel on the other side of it. The most informative rival is the one your guests actually compare you to, which isn’t always the one down the street. If you haven’t settled that, start with What’s Your Competitive Set When You Have No Brand Flag? and Building a Competitive Set from What Guests Actually Say.

2. Treating thin aspects as findings

A 14-point gap on 50 mentions looks dramatic and proves little. State the mention count next to every score, and when an aspect is thin, say so.

3. Comparing aspects only one hotel reports

Each hotel reports its top eight aspects. If an aspect is missing for one hotel, that doesn’t mean guests have no view on it. It means it wasn’t prominent enough to reach the top eight.

4. Reading the rating as the verdict

Across Lucerne hotels, rating and aspect-level sentiment move almost together, with a correlation of 0.97. That’s why the rating is a good summary, and also why it can’t tell you where the problem is. Two hotels can share a rating and have opposite strengths.

Themes are signals, not proof of cause, so treat whatever you find in the Staff reviews as hypotheses to test. For the Value for Money question, the next article in this series picks up the thread: next article in this series, “Is It Your Hotel or Your Market?”

Two real actions beat a ten-item list nobody owns. As we say in our other work, a strong average is a hiding place, not a finding. The same is true of a tie.