Estimates at design stage: what architects owe their clients

"What will it cost?" Every architect hears it while the set is still in progress — and the honest traditional answer has been some version of wait for the tender. Not because the question is unreasonable, but because quantities used to be expensive to produce: nobody re-counts a whole building every time a floor plan revision lands.

That scarcity shaped the whole workflow. Cost enters the conversation at the end, when the design is hardest to change. The owner discovers the budget problem after the drawings are finished, and value engineering becomes an exercise in undoing decisions instead of making them.

Design-stage quantities flip the sequence. If a takeoff can be re-run from the current drawing set in hours rather than days, then every issue of the set can carry its own quantity snapshot: how many square metres of glazing, how many doors of which type, what moved since the last issue. The conversation with the client changes from "we'll know at tender" to "here is what this revision did to the numbers."

And when the tender does come — often run by the architect on the owner's behalf — the same quantities become the levelling baseline. Bids get compared against an independent count, not against each other. A sub whose number is far off a traceable baseline has to explain the gap by pointing at drawings, which is exactly where that argument should live.

The prerequisite for all of this is trust in the counts. A number is only useful to a professional if it can answer one question: says who? That's why every quantity in a CAIRO takeoff carries its source — the sheet, the schedule row — and why a dimension the documents don't state shows up as an explicit gap instead of a silent guess.