how the benchmarks work
The receipts behind every verdict, and the things we'll never claim.
where the rates come from
real quotes + rate cards
Real Singapore renovation quotations (2024–2026) and published contractor rate cards.
cross-checked
Every item is verified against multiple sources before it enters the database. One quote is an anecdote, several are a rate.
stored as a range
Low · median · high, never a single "correct" price, because scope and materials legitimately move costs.
always in the trade's native unit:
how a verdict is decided
Verdicts are mechanical, not vibes. The AI only reads your quote and extracts the line items. It does not decide the rates, and it cannot cite a rate that isn't in our database. Here's a real example from the sample report:
kitchen cabinets, 28ft run, quoted S$4,480
S$4,480 ÷ 28ft = S$160/ft
Inside the range → fair , and the report says so. Fair pricing gets named too; that's what makes the flags believable.
inside the low–high range
over the high end, delta measured vs the median
more than 25% over the high end
no rate in our database: we say so, we never guess
your data
✓your quote file is used for analysis only, stored privately for the life of your project, deleted any time on request
✓anonymised line items improve the benchmark database, as set out in our terms
✓always anonymised, no names, contacts or addresses, ever
✓never sold or shared with contractors, IDs or anyone else
what this tool can't do
✕rate contractors. We audit quotations, never name or score firms
✕judge workmanship, a fair price says nothing about quality
✕capture every case, scope, materials and site conditions legitimately move prices
✕replace professional advice, informational benchmarking only
Landing statistics: budget-overrun figure from a Qanvast homeowner survey (600+ homeowners); complaint figures from CASE 2024 statistics on renovation contractors; ID vs direct-contractor spread observed across published rate cards and quotes in our database.