Keep a daily report that survives a claim

Open the site diary and create today from the previous day. The crews, areas and activities carry over, so the work is editing what changed rather than typing the whole day again: who is on which area, what got done, what stopped.

Why: A blank form at the end of a long day gets three words in it. A form that already reads like yesterday gets corrected, and a corrected form is a real record. This is the whole difference between a diary that is kept and one that exists in the contract only.

Put the day's hours in field time against the cost code they were worked on, and mark separately any hours that were time and materials rather than measured work. Where a crew split across two areas, split the hours rather than rounding them to whichever area was bigger.

Why: Hours booked to a cost code are hours you can defend later at a rate. Hours booked to a week are payroll and nothing else. When a change is priced on actual labor, this is the only record that tells you what the labor actually was.

For each machine on site record the hours it worked, the hours it stood idle and the hours it was down, rather than a single mark that it was present. Then record the weather with the thing it prevented: not just rain, but which pour was called off and which crew stood.

Why: Idle plant is a cost you can claim and down plant is a cost you cannot, and only the report written on the day can tell them apart. Weather recorded on its own proves the weather happened, which nobody disputes; weather recorded against the activity it stopped is what proves the delay.

Review the period as a run of days rather than as a list of reports, and give the empty ones an explicit reason: weekend, holiday, shutdown, weather, no access. A day with no work is still a day that gets a report saying so.

Why: Federal construction reporting asks for every calendar day to be accounted for, and a missing day is read as a record that was not kept rather than as a day nothing happened. The other side does not have to prove what was in the gap, only that the gap is there.

In claims evidence, collect the days that bear on one event into a single bundle with the photos, instructions and correspondence that belong to them, and put them in the order they happened.

Why: Evidence assembled while the job is running takes an hour. The same evidence assembled from an archive after the job closed takes a week and comes back incomplete, because the people who knew which photo belonged to which day have moved on.

Raise the variation from the bundle so the labor, plant and material on it come from the hours and equipment states already booked, rather than from an estimate made after the fact.

Why: A number that can be walked back to a specific day, a specific crew and a specific machine gets argued about on its merits. A number that cannot gets treated as a negotiating position, and negotiating positions get halved.