How it works
From ★ 4.9 to a score that means something
TrustBnB asks one question about every listing: compared with the other listings in its city, how good is this rating, once you account for how many reviews back it up? It answers in three steps. Try each one below.
Step 1
Weigh the reviews
A perfect 5.0 from three guests is mostly luck. A 4.95 from three hundred is a track record. TrustBnB blends each rating with the city's average rating, as if the listing had 10 extra reviews at that average. Listings with few reviews move a lot; listings with hundreds barely move.
Step 2
Rank it in its city
Next, TrustBnB works out the share of listings in the same city whose weighted rating is lower. Every listing goes through step 1 first, so they're all compared on equal terms.
That's why the city matters. Paris and Tokyo don't rate the same way, and a 4.85 that's ordinary in one city can be above average in another. TrustBnB has data on 44 cities, built from Inside Airbnb's public data, and recognizes the city from the page you're on. Everywhere else, it compares with all 44 cities together.
Step 3
Re-grade it on 1–5
Finally, the rank becomes a score on a scale people actually use from top to bottom. The default, Realistic, follows how TripAdvisor ratings are spread out: most places land around 4, and about 30% earn 4.5 or more. Strict is a straight ranking, where the median listing gets a 3.
Four tiers
Every score comes with a tier, so you can read it at a glance. On the Realistic curve, a Great listing is one that would show 4.5 stars or more on TripAdvisor.
| Tier | Scores | On the Realistic curve |
|---|---|---|
| 5.0 Great | 4.3–5.0 | 30% of listings |
| 4.2 Fine | 3.8–4.2 | 32% of listings |
| 3.7 Meh | 2.8–3.7 | 30% of listings |
| 2.7 Poor | 1.0–2.7 | 7% of listings |
What the score doesn't do
- It only looks at the overall rating and the number of reviews, not price, location, photos or what the reviews say.
- It compares a listing with its whole city, so a hostel bed and a villa share the same pool.
- It's as fresh as its data. Inside Airbnb collects each city every few months, and TrustBnB checks for new data monthly.
The formulas, the data pipeline and the code are all in the technical documentation.