How to Read a Hotel’s Review History Before Making an Offer

Key findings at a glance

• A rating can stay flat while one aspect slides and another rises. A buyer who reads only the headline number sees neither.
• Single hotels carry thin histories. In Interlaken (42 hotels, 12-month window), the median hotel has 346 reviews, about 29 a month. Cut that by month and by aspect and each slice is small.
• Take Maintenance & Upkeep. The average hotel reporting it has about 55 Review Mentions in Interlaken (1,040 across 19 hotels), about 75 in Lucerne (1,640 across 22) and about 62 in Lausanne (495 across 8). Split into quarters, that’s roughly 14 to 19 mentions a quarter, well under the 50-mention line we use to call an aspect readable.
• A seasonal comparison needs the same period a year earlier, so a buyer needs about two years of history, not one.
• The order matters: check the history can be trusted, then read direction, break, season and recurrence.

The rating is the same. The hotel isn’t.

Picture two hotels you’re thinking of buying. They’re hypothetical, but you’ve probably met their twins. Both show a 4.2 in the seller’s deck. Both have steady occupancy. Both look like safe assets.

The first climbed to 4.2 over two years, from 3.8. Recent guests describe a place that’s getting better. The second slid to 4.2 from 4.5. Recent guests describe a place that’s losing something, and the average hasn’t caught up yet.

Same number, different assets. One comes with momentum. The other comes with a question about what changed, and whether it’s still changing.

The seller’s deck gives you the rating. It rarely gives you the direction. That has to come from the history, and reading it takes more care than looking at a chart.

Why history matters more when you’re buying

A guest booking a room wants to know what tonight will be like. A buyer wants something different: what the hotel will be like in three years, and what it will cost to get there. That makes direction more valuable than level. A buyer inherits the trend, the backlog behind it and the reputation that has to be carried.

This article builds on two earlier pieces in the series. A Pre-Acquisition Guest Sentiment Checklist covers what to check. What Guest Reviews Reveal About a Hotel’s Deferred CapEx Before You Buy covers whose bill a weak upkeep score is. This one covers how to read the history itself.

Two terms before we go on. Review Mentions counts how many times guests mentioned an aspect. They aren’t unique reviews, because one review can touch several aspects. Hotels Reporting is how many hotels have the aspect among their top eight topics. It measures how prominent a topic is, not whether it’s absent elsewhere.

Read 0: Can you trust the history?

Before reading a trend, check that the history is a fair record. Five things can quietly distort it.

• Gaps and spikes. A quiet stretch followed by a burst of reviews can mean a review-request campaign, a new platform integration or a change in how guests are asked. A burst isn’t an improvement. It’s a change in who is writing.
• Platform mix. If the history draws more heavily from one platform in recent months, the trend may reflect the platform, not the hotel.
• Traveler mix. A shift from leisure guests to groups, or from domestic to international travelers, changes what guests value and complain about.
• Removals and edits. From the outside, removed or edited reviews are invisible. You can’t read what isn’t there, so say so in your notes.
• Changes inside the window. A new brand flag, a new operator or a closure for works changes what the hotel is. Mark the date before you read anything else.

The practical rule is simple. If the history can’t be trusted, say so and don’t draw a trend. A clearly flagged gap is more useful to the buyer than a confident line through bad data.

Read 1: Direction. Rising, flat or sliding?

Start with the shape of the line, using equal windows. Compare a rolling six or twelve months with the six or twelve months before it. Single months are too noisy to carry a conclusion, and a hotel with about 29 reviews a month can swing on a handful of guests.

Then look beneath the rating. A flat rating can hide one aspect sliding while another rises. Staff may improve while rooms age, and the two cancel out in the average. So read the rating first, then the aspects that carry the most Review Mentions, with counts printed next to each.

This is the series line in a new setting. A strong average is a hiding place, not a finding. So is a flat one.

What to look for: a direction that holds across at least three consecutive windows, on the same aspects. One down period is a month. Three is a pattern.

Read 2: The break. When did it change?

Trends have two shapes. Drift is gradual, with no single moment behind it. A break is a step: sentiment is one thing before a date and another after. Breaks are the more useful find, because they can be dated, and anything that can be dated can be asked about.

When you see a step, ask the seller what happened that month. A renovation, a new general manager, a brand or booking-channel change, a closure for works, a change in pricing: each can move how guests talk. Then check whether sentiment moved when the event did, or some time before it. A change that starts before the event it supposedly explains tells you the story is wrong.

A break after a renovation tests whether the capex worked. A break after a management change tests the operator. Either way, the date turns a vague unease into a specific question for the seller.

Caution: the date and the move can coincide by chance. A break is a lead to follow, not proof of cause.

Read 3: The season. Compare like with like.

Hotels have seasons, and a trend read across them can mislead in both directions. The fix is to compare a quarter with the same quarter a year earlier, not with the quarter before it. A hotel that dips from summer to autumn may be doing exactly what it did last year.

This is where data volume bites. A like-for-like comparison needs at least two years of history. The published SENTEEZ town benchmarks use a 12-month window, so they describe a level, not a seasonal trend. A buyer who wants the seasonal read has to ask for a longer history.

Seasonal shapes also differ by market type. Alpine, lake, city, coastal and village destinations have different peaks, different guest mixes and different review volumes, so a pattern in one doesn’t transfer to another. Compare a target with its own past first, and with its town second.

One seasonal trap deserves a mention. Guests in several of the towns we study connect cooling problems with upkeep, noise and value complaints. Those complaints are likely to cluster in warm months. A summer rise in them can look like decline when it’s a seasonal pattern. If the same rise shows up in the same quarter of the previous year, it’s the season. If it’s new, it’s the hotel.

Read 4: The recurrence. Do the same complaints keep coming back?

The last read separates a one-off from a pattern. A complaint that appears once, in one period, is an incident. A complaint that returns across periods is a condition.

What happens after a complaint tells you about the operator. If a theme appears, then fades and stays gone, the hotel probably fixed it. If it fades and returns, the fix didn’t hold, or the root cause is still there. If it never fades, nobody has found the cause.

Public management responses add a second layer. A response shows that a problem was acknowledged. It doesn’t show that it was fixed. The test is whether the same theme keeps appearing in later reviews after the response.

This is the time-based version of the previous article. In that piece, a weak Maintenance & Upkeep score was read against the town baseline and the town ceiling. Here the question is whether the same upkeep themes, such as cooling, bathroom and plumbing faults, aging rooms and slow fixes, return period after period at the same hotel. Recurring themes are signals, not proof of cause. They tell the engineer where to look first.

How much data do you need?

Less than a trend line implies, and more than a single hotel usually has. Our published datasets show why.

• Reviews per hotel are modest. Interlaken’s median hotel has 346 reviews in a 12-month window. Smaller markets run lower: medians of about 154 in Ticino, 184 in Carmel and 143 in St Ives.
• Aspect slices are smaller still. Each hotel reports its top eight aspects, and a review can touch several. The average hotel reporting Maintenance & Upkeep carries about 55 Review Mentions on it in Interlaken across the whole window, about 14 a quarter.
• The series rule applies. Aspects with fewer than 50 Review Mentions, or fewer than three hotels, are too thin to benchmark. For one hotel, we apply the same logic: under 50 Review Mentions in a period means no trend claim for that aspect.

In practice that means three habits:

• Use quarterly or rolling windows, not months.
• Print the count next to every figure.
• Flag anything thin and say what it can and can’t support.

A hotel with a thin history isn’t a bad hotel. It’s one where the honest answer to “is it improving?” may be “we can’t tell yet”, and that’s a finding worth putting in the file.

Three hypothetical targets

To make the reads concrete, here are three assets, each with two years of history split into six-month windows. They’re hypothetical, not real hotels, and the figures are chosen for illustration. Each upkeep window is assumed to carry about 60 Review Mentions, above the thin-data line.

Target A needs nothing from the history. The line is flat, the aspects are stable and each season matches the last. The only open question is level, and for that the town baseline and ceiling from the previous article do the work. A clean history is a finding too, and it shortens the questions you take to the seller.

Target B carries a dated break. The rating holds for a year, then drops as upkeep falls from the mid-thirties to 14%, and Staff would be worth checking for the same step. If the seller says a new general manager arrived at that point, the history gives you a specific question: what changed in how the hotel is run, and what has been done since? Ask for the event log, the maintenance and work-order records for the same months, and the capital plan. Because the step has a date, it goes into the offer as a hotel-specific item.

Target C is the one a headline number hides. Rating and upkeep sit exactly at their long-run levels, so a year-on-year check passes it. The recurrence read says otherwise: cooling takes a bigger share of summer complaints in each of three summers. A complaint that returns and grows is a root cause nobody has fixed, and it may be one capital item behind several scores. Ask what was done about cooling after each summer, when the units were installed and what the next plan is. Treat it as a management question, not a line item.

What to do with the finding

A weak trend isn’t a reason to walk. It’s a reason to ask a sharper question, request a specific record or build a line item into your offer, and all three are easier before signing than after.

The disclosures worth asking for:

• A dated event log: renovations, management changes, brand or channel changes and closures, with the month each happened.
• The seller’s own guest-satisfaction data for the same period, to compare with the public history.
• Complaint, maintenance and work-order logs matched to the months where the trend moved.
• Installation dates and service records for cooling, plumbing and electrical systems, where recurring themes point to them.
• A longer review history, at least two years, if the seasonal read can’t be made from what you have.

Then price it by type. A market pattern goes into the capex budget. A dated, hotel-specific break goes into the offer. A recurrence that persists across periods goes to the seller as a management question.

What this analysis cannot tell you

Four limits are worth stating plainly.

• Reviews describe what guests say, not what’s inside the hotel. They show where guests keep running into friction. They don’t show roofs, boilers, structure or wiring until something fails.
• Trends are associations, not causes. A drop after a management change doesn’t prove the change caused it, and a rise after a renovation doesn’t prove the renovation worked. Themes are signals, not proof.
• Coverage is uneven. Aspects with fewer than 50 Review Mentions or fewer than three hotels are too thin to benchmark, and a single hotel’s history is often thinner than that. Hotels Reporting also measures prominence, so a hotel without an aspect in its top eight topics isn’t automatically clean.
• This isn’t a survey. It sits next to financials, a property condition assessment and legal diligence. It replaces none of them, and the thresholds in this piece are working guidelines, not tests.

See your own version of this analysis
The free SENTEEZ teaser report compares a hotel with one competing hotel, shows where each wins and loses, and lists the top recurring guest themes. It can be requested for an acquisition target and is delivered within 48 hours, with no commitment.
For acquisitions, Hotel Acquisition Due Diligence goes deeper: trend history, the full aspect breakdown and benchmarking against comparable hotels in the same market. It’s part of the Enterprise plan.
Get your free teaser report →

Three steps before you make an offer

A review trend is worth acting on only when it holds across several periods, shows up on the same aspects, and survives a like-for-like seasonal comparison.

• Check the history is fair. Look for gaps, spikes, platform or traveler shifts and dated changes. If it can’t be trusted, say so and don’t draw a trend.
• Read direction, break, season and recurrence, in that order. Use equal windows, date every step, compare the same quarter a year earlier, and print the count next to every figure.
• Decide: ask, adjust or accept. A market pattern goes in the capex budget. A dated, hotel-specific break goes in the offer. A recurrence that persists goes to the seller as a management question.

A strong average is a hiding place, not a finding. So is a flat one.