A close-out captures what was actually there against what was sold, on the day it finishes, written by somebody who was on site. Its job is not to grade the crew. Its job is to tell the estimate what it got wrong.
Written by JD Davenport, who still sells for a living. Updated 4 November 2026.
Most close-out forms are a quality check with a photo requirement bolted on. They tell you the job looks finished. They tell you nothing about whether it was worth doing, and that is the question you will be asking in three months when the margin comes in low and nobody can say why.
A close-out written on Monday about a Thursday job is a guess. People do not remember how many extra layers were on a roof four days later, and the number they write down will be round, and round numbers are the ones that do not reconcile.
So: the day it finishes, by the crew leader or the supervisor who was standing there. Not the office, from the file.
The paperwork, with a reason when it is absent. The labor price sheet complete and signed by both parties, photographed, attached to the job. Receipts attached. And a field for why, if no price sheet was done. Missing paperwork is not fraud, and a blank you cannot explain later is worse than a blank with a sentence next to it.
What was actually there. This is the part most forms skip and the part that pays for the whole exercise. The real quantities. For a roof: the pitches and the size of each, decking sheets replaced, dimensional lumber and its size, extra layers found beyond the estimate and how many squares of them, bundles used, chimneys and their size, skylights, and anything the estimate did not carry at all. Your trade will have its own nine.
The quality gates. Substrate inspected after tear-off, with any change told to the customer and to production before the next stage and photographed. The weather barrier verified before the covering went on. Fastening and flashing checked against a written standard. Surface protection everywhere access equipment touched the building.
Did anything go wrong, and unplanned spend. Two separate questions. Things go wrong on jobs that cost nothing, and things cost money on jobs where nothing went wrong.
Was it sold right. One field, and it is the whole point.
Nobody is scored by a close-out except the estimate.
That sentence has to be true in practice or the form stops being accurate within a month. The moment a crew learns that writing down four extra layers gets somebody a conversation, the extra layers stop being written down. Then your estimating data is clean, agreeable and wrong, and you have traded the one instrument that could have told you what work actually costs for a quiet Thursday.
If you want the data, you have to make it safe to record. That is a design constraint, not a nicety.
On its own, nothing. A close-out that gets filed is a diary.
The use is the loop: what the close-out catches this month becomes a line on the pre-install review next month. An extra layer found on three jobs in a row is not three surprises, it is a question the estimate should have been asking all along. Written down, it becomes one. Left in a folder, it stays three surprises and then a fourth.
That loop is the difference between a company that has a close-out form and a company whose estimates get better every quarter without anybody being told to try harder.
The close-out gate, the pre-install gate it feeds, the record between them and the board that runs both are The Production Standard. What the quantities on it are worth, per job, is The Job Cost Standard. More answers on the questions page.