The site diary template,
usable straight off this page.

This page sets out what a daily site diary entry captures, in the order a supervisor would write it at the end of the day. It is deliberately a field list rather than a downloadable form, because the diary's failure mode has never been the layout. It is the empty diary, the one written up on Fridays from memory, and the one kept only on days that felt important. The fields hold on any job; the discipline section below is what makes them worth keeping.

Why the record carries the weight it does is covered in the site diaries and daily records reference, the claim machinery the diary feeds is in extensions of time and delay, and the long-tail value of the photo record is in records as warranty defence. Everything here is general information, not legal advice.

Eight fields, the same ones every day. The test of an entry is whether someone reading it in three years, with no memory of the job, could reconstruct the day from it.

• Date, job and author. Who wrote the entry and when. An unsigned, undated note is an anecdote. The author matters because the entry is that person's account of a day they were present for, which is what gives it standing.
• Weather, and what it did to the work. Morning and afternoon conditions, and the effect: whether trades could proceed, what stopped, who was stood down. "Rain" is a fact about the sky; "site closed, slab pour cancelled, concreter stood down" is a fact about the job, and only the second one supports a claim.
• Trades on site. Which companies, how many people, what they worked on. This is what later shows a stage was progressing, a trade was delayed by others, or a disruption cost real labour, rather than asserting it without support.
• Deliveries and materials. What arrived, what was short or damaged, and what failed to turn up. A late delivery that stalls a trade is a programme event, captured while the docket and the memory are both fresh.
• Delays and disruptions. Anything that stopped or slowed work, and why: the trade that did not show, the failed inspection, the access problem, the design query holding a stage. These entries are what connect a cause to a programme impact when a delay has to be justified.
• Instructions and decisions. Anything directed, agreed or decided on site, and by whom. The change directed verbally and never recorded is the classic source of a disputed variation, and this field is where the conversation is captured before it becomes a disagreement.
• Inspections and visitors. Who came to site, what they looked at and what resulted: the certifier, the engineer, the client walking their own build. The inspection that passed, and what it covered, is evidence the job will want later.
• Progress and photos. What was completed, tied to dated photographs, including what is about to be covered up. The photo of what was under the slab, taken before the pour, settles arguments nothing else can.

The common version is not the builder who keeps no diary. It is the builder who keeps one for three weeks, then gets busy, then catches up on Fridays, then stops, and only discovers the gap when a dispute reaches back for a fortnight in June that nobody recorded. The wet days are in nobody's record, the trades who were stood down are a guess, and the claim that should have been routine paperwork becomes an argument about whose memory to prefer. The diary was abandoned precisely because nothing was going wrong at the time, which is the only time a diary can be built.

The second version is subtler: the diary kept only on the bad days. It reads as selective because it is, and a record that only ever describes problems invites the question of what the ordinary days would have shown. Completeness is not a nice-to-have; it is what makes any single entry believable. The five-minute entry on an uneventful Tuesday is what buys credibility for the entry on the day everything went wrong.