Every opening measured before anything is ordered
Here is the thing nobody mentions when you are looking at window brochures: in an older house, no two openings are the same size. Not approximately the same, not the same within a tolerance you can ignore. The house was framed by people working to a standard that was looser than today's, and then it settled for several decades. Somebody re-headered an opening for a bigger unit in the seventies. Somebody else replaced three windows on the sunny side and left the rest. A wall took on water, dried out and moved. By the time anyone is quoting you, the openings on one elevation can differ by a half inch from each other, and an individual opening can be a half inch out of square corner to corner.
Which is why the useful measurement is not one measurement. Each opening gets measured for width at the top, the middle and the bottom, for height at both jambs and up the centre, and diagonally both ways to find out how far from square it actually is. The controlling dimension is the smallest one, because that is the one the window has to pass. Then it gets written down, opening by opening, and the order is built from that list rather than from a room count.
The reason that is worth doing is not fussiness for its own sake. It is that custom-sized units from a manufacturer that builds to order cost close to what standard sizes cost. The window is being made either way; the machine does not care whether the number ends in a whole inch. So measuring properly buys you a window that actually fits the hole in your wall, at very nearly the price of one that does not — which is an unusually good trade, and the only reason to skip it is that measuring twenty-odd openings takes a person the better part of a morning.
The alternative is what happens when someone orders to the nearest stock size down and makes up the difference on site. That difference does not disappear; it gets absorbed. It becomes a wider gap around the frame, packed with shims and foam, and then covered with a broader trim or an extension jamb so you cannot see it. Four things follow from that, and a homeowner feels all of them eventually. You lose glass — a window ordered two inches small in each direction is a permanently smaller view and a permanently darker room. The unit sits in a cavity that is harder to insulate evenly and easier to leave with a void in it, which is where the draught you cannot find comes from. The frame is being held by fasteners and shims rather than bearing on the opening, and over years of seasonal movement that is what makes a sash bind, a lock stop lining up and a seal stop sealing. And the trim gets fatter, which is the tell — if the replacement windows in a house have noticeably wider casing than the originals, something was being covered up.
None of this shows up on the day the job finishes. Everything works, everything looks clean, and the problems arrive in the third or fourth winter, in the form of one window that fogs, one that sticks, one room that is colder than the rest.
So the questions to ask whoever is quoting your house are short and they are diagnostic. Does someone field-measure every opening, or is the quote built from a count of windows? Are the units made to those measurements? And what happens at the openings that are not square — because in a house of any age there will be some, and the answer to that question tells you most of what you need to know about the rest of the job.
