Booking.com Review Score: How It's Calculated and What a Good Score Really Is
This guide explains how the Booking.com review score is calculated, what each label means, and what counts as good, using live scores from 2,523 properties in three cities.
Table of contents
Your Booking.com review score says 8.4, the label says "Very Good", and your team can't tell whether that's strong.
The score itself is simple to describe.
It's a recency-weighted average of the 1 to 10 overall ratings guests gave in the last 36 months.
The subscores shown beside it stay out of the math.
Whether 8.4 is good depends on the city.
We pulled 2,523 live scores across Lisbon, London, and New York, and the same number lands in very different places in each.
This guide covers the calculation, the label scale, a worked example, and the benchmarks.
Key Takeaways
- Recent reviews count more since January 2025. Booking.com weights the overall score by recency, so the newest review moves it most and older reviews fade until they're archived at 36 months.
- Subscores don't feed the overall score. At all 120 properties we checked, the overall sat below the average of the six subscores, by a median of 0.34 points.
- "Good" is a bottom-third score. Between 33% and 38% of properties in each city scored below 8.0, and the median property was "Very Good" (8.2 to 8.4).
- A 9.0 means different things by city. It put a property in the top 23% in Lisbon, the top 13% in London, and the top 9% in New York.
- Label words change by language; cutoffs stay put. US English shows "Wonderful" and "Excellent" where UK English shows "Superb" and "Fabulous" for the same numbers.
- Benchmark against your comp set. Your own history sits in the Extranet, and competitor scores can be pulled in bulk through a Booking.com data API.
How Is the Booking.com Review Score Calculated?
Booking.com asks every guest who booked a stay one required question: an overall score from 1 to 10.
That rating is the only input to the number on the property page.
Guests get an email after checkout and have 90 days to leave a review, according to Booking.com's guide to Guest Review Scores.
The score covers reviews from the last three years.
Reviews stay on the page for 36 months, then get archived, and the score is recalculated from what's left.
Since January 2025, the average is weighted by recency.
In Booking.com's words, "the most recent review has the biggest impact," and "as a review gets older, its importance decreases."
For properties with only a few reviews, Booking.com says it limits big swings by weighting the most recent reviews equally.
Its explainer video adds two effects: recent improvements show up quickly, and low scores from the off-season weigh less.
Booking.com doesn't publish the weights themselves.
| Input | In the overall score? | Detail |
|---|---|---|
| Overall rating, 1 to 10 | Yes | The only required question; one number per guest |
| Recency | Yes, since January 2025 | The newest review weighs most; older ones fade |
| Reviews older than 36 months | No | Archived, and the score is recalculated without them |
| Six subscores | No | Cleanliness, comfort, value, facilities, location, staff |
| Wi-Fi and breakfast ratings | No | Optional extras |
| Review text | No | Read by guests, never averaged |
This setup dates from a 2019 to 2020 overhaul.
Before it, Booking.com used a 2.5 to 10 scale and built each score from the average of six items, per a 2021 study in Tourism Management.
So a score from 2018 and a score from today don't sit on the same scale.
What Moves the Score Now
Under a plain average, a property with 3,000 reviews barely twitches when a new one arrives.
Under recency weighting, a run of good or bad recent stays shows up faster, in both directions.
You can see this from the outside.
For 120 properties, we compared the displayed score with the plain average of each property's own 1 to 10 review histogram.
They matched to one decimal at 33 properties (28%).
At 38, the displayed score sat at least 0.1 above the plain average; at 19, at least 0.1 below it.
Hotel 48LEX New York shows 8.1 while its 400 reviews average 7.77, and one London property shows 5.5 against a histogram average of 6.16.
That gap is most likely the recency weighting at work: recent stays that beat a property's history lift the shown score, and worse ones drag it down.
Why the Subscores Don't Add Up to the Overall
Next to the overall score, Booking.com shows subscores for cleanliness, comfort, value, facilities, location, staff, and Wi-Fi.
It's natural to read the overall as their average.
In our data, it sat below that average every time.
We pulled the full score breakdown for 120 properties, 40 per city, each with at least 200 reviews.
The overall sat below the average of the six main subscores at all 120, by a median of 0.34 points and by as much as 1.08.
At 18 of them (15%), the overall was lower than all six.
Room Lisboa, with 14,155 reviews, is a clean example:
| Score | Room Lisboa |
|---|---|
| Overall | 8.1 |
| Location | 9.2 |
| Staff | 8.8 |
| Wi-Fi | 8.7 |
| Value for money | 8.5 |
| Cleanliness | 8.3 |
| Comfort | 8.3 |
| Facilities | 8.2 |
The reason is in Booking.com's guest review guidelines: guests submit subscores and the overall score independently, "so there's no direct link between them."
A guest can give location a 10 and the stay a 7 because of one loud night.
Location was the top subscore at 82 of the 120 properties (68%), which lifts the subscore average for anything central.
Booking.com says it leaves subscores out of the overall partly because some, "such as property location," sit outside a property's control.
For a hotelier, the overall is the ranking number and the subscores are the diagnosis.
When one subscore lags, the review text behind it explains why; our guide to hotel review sentiment analysis covers how to read that text at scale.
Booking.com Review Score Calculator: A Worked Example
Booking.com doesn't offer a review score calculator, and it doesn't publish its recency weights.
You can still check the plain math from a property's score histogram.
Here's Room Lisboa's, pulled on October 6, 2026:
| Guest rating | Reviews |
|---|---|
| 10 | 4,713 |
| 9 | 2,349 |
| 8 | 3,093 |
| 7 | 1,690 |
| 6 | 729 |
| 5 | 537 |
| 4 | 296 |
| 3 | 195 |
| 2 | 127 |
| 1 | 426 |
Multiply each rating by its count and the products add up to 114,353.
Divide by 14,155 reviews and you get 8.08, which Booking.com shows as 8.1.
One guest in three (33.3%) gave a perfect 10, while 16.3% gave a 6 or lower.
histogram = {10: 4713, 9: 2349, 8: 3093, 7: 1690, 6: 729,
5: 537, 4: 296, 3: 195, 2: 127, 1: 426}
total = sum(histogram.values())
plain_mean = sum(score * count for score, count in histogram.items()) / total
print(total, round(plain_mean, 2)) # 14155 8.08
The recency weighting is where a calculator gets hard.
Here's an illustration with made-up numbers, since Booking.com's real weights aren't public.
A property has 40 reviews: 30 older ones averaging 7.5 and 10 recent ones averaging 9.0.
The plain average is 7.9.
Any weighting that favors recent reviews pushes the shown score above 7.9, and the stronger the weighting, the closer it gets to 9.0.
That's also why a property recovering from a bad year sees its score move faster now than it did before 2025.
Booking.com Rating Scale: What Each Label Means
Every property score gets a word next to it.
Booking.com doesn't publish the cutoffs, so we read them off the 2,523 scored properties in our US English pull, then checked the UK English words on the same 1,098 London properties.
| Score | US English label | UK English label |
|---|---|---|
| 9.5 to 10 | Exceptional | Exceptional |
| 9.0 to 9.4 | Wonderful | Superb |
| 8.6 to 8.9 | Excellent | Fabulous |
| 8.0 to 8.5 | Very Good | Very good |
| 7.0 to 7.9 | Good | Good |
| 6.0 to 6.9 | Pleasant | Pleasant |
| 5.6 to 5.9 | Okay | Okay |
| 5.0 to 5.5 | Fair | Passable |
| 4.0 to 4.9 | Disappointing | Disappointing |
| Below 4.0 | Poor (Bad at the very bottom) | Poor (Bad at the very bottom) |
The bottom rows are thinly sampled, because few properties live there.
We only saw "Bad" on scores of 1.0.
Single reviews use a different set of buckets.
Booking.com's review data groups individual ratings as Wonderful (9 to 10), Good (7 to 8), Okay (5 to 6), Poor (3 to 4), and Very Poor (1 to 2).
So one guest's 8 counts as "Good," while a property averaging 8.0 is "Very Good."
One subscore does gate something.
For homes and apartment-style properties, Booking.com's quality rating program requires an overall cleanliness score of 6.0 or higher once a property has a review score.
What Is a Good Booking.com Review Score?
We searched Booking.com for the same two-night stay (November 12 to 14, 2026, two adults) in three cities and kept every property that had a score.
| Lisbon | London | New York | |
|---|---|---|---|
| Properties with a score | 953 | 1,088 | 482 |
| Median score | 8.4 | 8.2 | 8.2 |
| Middle half of properties | 7.7 to 8.9 | 7.6 to 8.6 | 7.8 to 8.6 |
| Share at 9.0 or higher | 22.7% | 13.1% | 9.3% |
| Share below 8.0 | 32.7% | 38.4% | 34.2% |
| Share below 7.0 | 9.3% | 11.4% | 7.3% |
| Median, 4-star properties | 8.8 | 8.4 | 8.3 |
| Median, 3-star properties | 7.9 | 7.8 | 8.0 |
Three things stand out.
The median property was "Very Good" in all three cities.
"Good" (7.0 to 7.9) is a bottom-third score: a third or more of properties scored below 8.0 everywhere.
And a 9.0 is common in Lisbon and rare in New York, top 23% in one and top 9% in the other.
Star class moves the bar too.
In Lisbon, the median 4-star property scored 8.8 and the median 3-star 7.9, so compare yourself within your class.
Booking.com's own bar is useful context.
Its Traveller Review Awards 2026 went to properties with an average between 8 and 10 and at least three reviews, measured on December 1, 2025.
Roughly two in three properties in each city sat at 8.0 or higher.
One caveat on the sample.
Booking.com serves about 1,100 results per search, so Lisbon (2,841 matches) and London (about 5,550) are the first 1,100 in its default order, which leans toward bookable, well-ranked properties.
New York is every one of the 497 properties available for those dates.
For a cross-check on a different platform and scale, our London review score benchmark tracks TripAdvisor ratings for 60 properties.
Tracking Your Score Against Your Comp Set
Your own score history is free.
In the Extranet, the Guest reviews tab shows how your overall score and category scores have changed over time.
The gap is everyone else.
Competitor scores are public on each property page, but checking 20 of them every week by hand gets old fast.
| Need | Extranet | Booking.com pages by hand | Data API (StayAPI) |
|---|---|---|---|
| Your own score trend | Yes, built in | Today's number only | Today's number; you keep the history |
| Competitor overall scores | No | Yes, one page at a time | Yes, one call per property |
| Competitor subscores and histogram | No | Subscores only, one page at a time | All seven subscores plus the 1 to 10 histogram |
| 20 competitors, weekly | Not possible | About 20 page visits a week | One scheduled script |
| Cost | Free | Free | Free tier, then a paid plan |
Hotel data APIs read the public score and return it as JSON.
StayAPI's Booking.com review scores endpoint takes a Booking.com hotel ID and returns the overall score, seven subscores, the histogram, and the review buckets:
curl -G "https://api.stayapi.com/v1/booking/hotel/reviews/scores" \
-H "x-api-key: YOUR_API_KEY" \
-d hotel_id=3590447
The response, truncated:
{
"success": true,
"hotel_id": "3590447",
"data": {
"scores": {
"overall": 8.1,
"location": 9.2,
"staff": 8.8,
"cleanliness": 8.3,
"comfort": 8.3,
"facilities": 8.2,
"value_for_money": 8.5,
"wifi": 8.7
},
"total_reviews": 14155,
"score_distribution": [
{ "score": 10, "count": 4713, "percent": "33.30" },
{ "score": 9, "count": 2349, "percent": 16.59 }
],
"score_percentage": [
{ "score_word": "Wonderful", "score_start": 9, "score_end": 10.1, "count": 7062, "percent": 50 }
]
}
}
One tradeoff: the API reads the public score, so it can't show the private guest-experience data your Extranet holds.
To try it on one hotel without code, the Booking.com reviews summary tool runs the same lookup in a browser.
If you need the review text as well as the scores, our guide to the Booking.com reviews API covers the access routes.
Getting Started
- Open the Extranet's Guest reviews tab and note your current score and your weakest category.
- List 5 to 10 comp-set properties in your city and star class, and grab their Booking.com hotel IDs; our Booking.com hotel ID guide shows where to find them.
- Sign up at stayapi.com (free tier, 50 requests, no credit card) and pull each competitor's scores.
- Schedule a weekly pull with the Booking.com reviews to Google Sheets integration, or build the same flow in N8N or Make.com.
- Compare against the comp-set median and your star class, and watch for competitors whose scores jump after a renovation or slide after a bad season. For portfolio-scale monitoring across many properties, book a demo call.
FAQ
How is the Booking.com review score calculated?
It's a weighted average of the 1 to 10 overall ratings guests gave in the last 36 months.
Since January 2025, Booking.com gives recent reviews more weight and lets older ones fade.
Subscores, Wi-Fi and breakfast ratings, and review text don't count.
What is a good Booking.com review score?
In our three-city sample, the median property scored 8.2 to 8.4 ("Very Good"), and a third or more scored below 8.0.
A score of 8.5 or higher beat most properties in all three cities, and 9.0 put a property in the top 9% to 23%.
Compare within your city and star class before reading anything into the label.
Is there a Booking.com review score calculator?
Booking.com doesn't offer one, and it doesn't publish its recency weights.
You can compute the plain average from a property's 1 to 10 histogram, as in the worked example above, and expect the shown score to drift from it when recent reviews differ from older ones.
Do subscores affect the overall Booking.com score?
Only the overall rating counts toward the score.
Guests rate the subscores separately, and Booking.com leaves them out of the calculation.
In our sample of 120 properties, the overall always sat below the average of the six main subscores.
What does "Wonderful" or "Superb" mean on Booking.com?
Both mean a score from 9.0 to 9.4.
US English pages say "Wonderful" and UK English pages say "Superb" for the same range, just as "Excellent" and "Fabulous" both cover 8.6 to 8.9.
How long do reviews count toward a Booking.com score?
For 36 months.
After that, reviews are archived and the score is recalculated from the reviews still online.
Under recency weighting, an older review counts for less well before it's archived.
Are booking.com reviews accurate?
Booking.com only takes reviews from guests who booked through its platform, and its guidelines say it may only accept them within 3 months of checkout.
It uses people and automated systems to detect fake reviews and deletes the ones it finds.
The score still reflects who chose to leave a review, so a large review count tells you more than a single number does.
Ready to simplify your hotel data?
Join other developers using StayAPI to build the next generation of travel applications. Get started for free today.