Answers

What do adjudicators look for in site records?

Records made at the time rather than assembled for the claim, covering every day rather than only the disputed ones, consistent enough that a pattern can be counted, and specific about what happened, when, and what it affected. Volume is not the test. An adjudicator working to a statutory timetable is looking for records that make the point quickly and can be relied on.

Last reviewed 20 Nov 2026

Security of payment adjudication in Australia runs to short statutory timeframes. An adjudicator typically has around ten business days to decide, and will be reading two accounts of the same project written by people who disagree. What survives that reading is not the longest submission.

This is general information rather than legal advice, and the requirements differ between jurisdictions. Check your own contract and the Act that applies to you.

Made at the time, not for the claim

A bundle assembled after a dispute started invites an obvious question: what was being recorded before anybody knew there would be an argument? Records that exist for every day, kept the same way whether or not anything went wrong, answer it without having to make the point.

The corollary is uncomfortable. Records that start being kept properly the week a dispute begins still help with that week, but they also draw attention to the gap behind them, and the gap gets read as something too.

Complete, including the uneventful days

Gaps get read as something. A diary with entries on the days that support a claim and nothing on the days around them is a diary somebody curated, whatever the truth of it.

Covering every day is not about diligence for its own sake. It is what makes the eventful days credible.

Consistent enough to be counted

A claim that says a particular cause recurred is far stronger when the records can show how often, across which weeks, affecting which trades. That requires the same cause to have been recorded the same way each time, which in turn requires a fixed list rather than free text.

It is the difference between asserting a pattern and demonstrating one.

Specific about effect, not just event

That a delivery arrived late is an event. That it arrived late, that a named task could not start because of it, and that the task sat on the chain that moved the completion date, is an argument.

Linking the cause to the work it stopped, on the day, is the part most records leave to be reconstructed afterwards, and reconstruction is exactly what the other side will attack.

Capable of being relied on

A document that could have been altered at any point since is weaker than one that could not. This is why bound and page-numbered diaries have been used for a century, and why the same question gets asked of a spreadsheet: when was this written, and how would anybody know if it had changed.

Whatever system is used, the answer to that question should be something other than "you would have to take our word for it".

See it on a real record.

Walk around a project with five weeks of records on it, or download a blank Day Record and keep it yourself. Neither needs a sign up.