Nobody's favourite restaurant is the highest rated one. Choose across twelve pairs of rooms and ReviewU maps what U actually like, then hands U the ones that match it.
All rooms, tuned to the signature U just built. Filter by cuisine, price, occasion and style. Never sorted by stars.
Two real rooms. Which one tonight? Twelve times.
Each pair isolates one axis, so twelve answers locate all seven.
U get a silhouette and a name, against the pool average.
Never by crowd average, star count or who paid.
Every site shows a price level. Almost none says what it means. Ours is one line, the same for all rooms.
Representative means the median dish. Taking the range end to end is how a pizzeria lands in the same bracket as a tasting-menu room. The £67 beef rib is real. It is not what most people order.
of rooms checked against a source, still weak. We refuse discount platforms, which show a cut price, and delivery apps, which mark it up. Where the format does not fit, the card says so.
A star average collapses everyone's taste into one number, then ranks by it. That rewards whatever suits the median visitor, not what fits U. ReviewU never asks for or shows one.
The pool is built to spread deliberately across cuisines rather than let a handful of famous categories dominate, and within each cuisine, to include genuinely different kinds of places, not just the most written-about one.
A cuisine that only shows up at one price point isn't a real choice: it's a filter with one setting. Prices are shown in £ bands built from real menus and reviews, dated and sourced, never estimated.
Every price, award and detail on a room's card is sourced: an official menu, a dated review, a certified listing. If something can't be confirmed, it's left blank rather than guessed at.
No Google rating, no review count, no aggregate score ever enters the pool or the matching engine. Fit to U's taste is the only ranking that exists here.
London only, for now.