How the DutchClassFinder score works
Every scored school gets a number from 0 to 10. It is our independent judgment, computed the same way for every school from public data. It is never edited by hand, and never influenced by payment. Schools cannot buy a better score, and 145 of 252 schools currently qualify for one.
The three components
Reviews56.25%
The school's Google rating, adjusted for how many reviews stand behind it. A 5.0 from 8 reviews scores lower than a 4.8 from 400, because a handful of reviews is a weaker signal than hundreds. Schools with fewer than 5 reviews are not scored at all.
A typical school: a 4.5 rating from 50 reviews scores 7.3.
What a 10 takes: a near-perfect rating backed by a large number of reviews. A 4.9 from 177 reviews scores 9.2; a 5.0 from 500 reviews scores 9.9. A flat 10 is out of reach by design: there is always room for more evidence.
Course range31.25%
How much of the Dutch-learning journey the school covers: CEFR levels taught (A0 through C1), the variety of course formats offered, and whether it prepares students for the official exams (Staatsexamen NT2 or inburgering). This course-offer data is verified against each school's own website (re-checked yearly); each school page shows the date we checked, and schools we could not verify say so.
A typical school: four levels, two formats, and exam prep scores 6.8.
What a 10 takes: all six levels from A0 to C1 (worth 5 points), four or more course formats such as group, private, intensive, and online (3 points), and official exam preparation (2 points).
Trust & transparency12.5%
What we could independently verify. The biggest part is independent quality oversight: whether the school has voluntarily submitted to a third party's audit that we can check in a public register, either the Blik op Werk keurmerk (required for DUO funding) or CRKBO registration (the audited register for private education). One is enough: both attest the same thing, an outside body has looked and vouches, and requiring a specific one would punish schools it doesn't apply to. The rest: prices published openly on the school's website, a publicly rated Google Business listing (re-checked monthly), and a full address and phone number. Not appearing in a register means "unverified", never "not registered". Social media presence is deliberately not counted.
A typical school: a school with no register match but published prices, a Google listing, an address, and a phone number scores 6.0.
What a 10 takes: register verification (4 points) plus published prices (2), a rated Google listing (2), a full address (1), and a phone number (1).
The headline score is the weighted average of the three components. We never guess at missing data and never penalize a school for information we haven't collected yet.
Why not just show the Google rating?
We do show it, on every listing. But most language schools have excellent Google ratings, so a wall of near-perfect stars tells you very little about which school fits better. The score adds what the stars leave out: how much evidence backs the rating, how broad the school's offering is, and what we could independently verify.
Why prices don't move the score
We collect and display verified course prices on every school page, with the date we checked them and a link to the source. But we deliberately don't grade schools on being cheap: a pricier course with smaller classes or more contact hours can easily be the better deal, and that judgment is yours to make. What the score does reward, under Trust & transparency, is publishing prices openly at all.
What the labels mean
- 8.5+ Excellent: outstanding on nearly every measure
- 7.0+ Very good: a strong choice, well above average
- 5.5+ Good: solid, with some weaker areas
- below 5.5 Fair: mixed reviews or a narrow offering
A "Fair" score is not a warning sign. It usually means a school that is new, small, specialized, or still collecting its first reviews. And "not yet scored" means exactly that: too few reviews to judge fairly, not a hidden bad score.
Known limitations
No formula is neutral, so here is where ours is weakest. Course range rewards breadth, which means an outstanding school that deliberately teaches only one thing (say, NT2 exam prep) scores lower there than a broad generalist; the score can't yet tell a narrow focus from a narrow offering. The Reviews component leans on Google ratings because they are the only quality signal available for every school; that inherits Google's blind spots, including that a rating earned years ago counts the same as one earned last month. And register verification depends on the registers themselves being current; we re-check them periodically, not live.
We'd rather name these trade-offs than pretend the number is perfect. When one of them can be fixed with better data, we fix it: we now log every school's rating history monthly, and a future version may use it so that fresh, active review bases count for more than stale ones.
Versioning and corrections
This is formula version 1.4. When the formula changes, we update this page and the version number. Most recently, v1.4 retired a price-based "Value" component: it quietly rewarded cheapness and could rank a school that hides its prices above one that publishes them, which is the opposite of what we stand for. Ratings and review counts are refreshed from Google monthly.
Spot something wrong about a school? Tell us and we will check it.