Records that survive
Adjudication compresses a year of a project into twenty-eight days of reading. What you can prove is what you wrote down at the time, and nothing you write afterwards improves it.
QScope Team · 28 March 2026 · 6 min read
Adjudication is a documents exercise. There is rarely a hearing, rarely live evidence, and never enough time. An adjudicator reads two bundles and decides.
Under that constraint, the record made at the time is worth several times the argument made afterwards. Understanding why changes what a surveyor bothers to write down.
What contemporaneous actually means
Made at or near the time of the event, in the ordinary course of business, by someone who knew what happened.
Three consequences follow.
Timing is visible. A record created in March about a March event carries weight. The same content produced in November carries much less, and it is usually apparent which is which.
Ordinary course matters. A site diary kept every day, including on the days nothing happened, is credible. One that exists only for the disputed fortnight invites the obvious question.
Consistency is tested. A record that contradicts your other records damages both. Adjudicators read the bundle as a whole, and the internal inconsistencies are what they notice.
The five records that decide most disputes
1. The instruction
What was instructed, by whom, on what date, in what form. Most variation disputes are entitlement disputes, and entitlement is decided here.
2. The notice, and proof of service
The payment notice or pay less notice, its date, and evidence it was sent to the right person in the right way. The document alone proves it was written, not that it was served.
3. Resource allocation
Who was on site, on what activity, for how long. This is what makes disruption and dayworks claims capable of assessment. It is also the record most often absent.
4. Progress at a point in time
Photographs with dates, marked-up programmes, progress records. Delay claims turn on what was actually built by when, and memory does not settle it.
5. The decision and its reason
Why a percentage was reduced, why a rule was chosen, why a claim was assessed at a particular figure. The reason is what makes the figure defensible, and it is the piece almost never recorded.
Records that do not survive
- Undated notes. A record without a date has no place in a chronology
- Files edited after the event. A spreadsheet whose figures have been changed with no history is a current statement, not a record of what was thought at the time
- Documents only one party ever saw. An internal note may be honest and it is not evidence of what was communicated
- Summaries without underlying data. A schedule of hours with no allocation sheets behind it is an assertion in a table
- Chains where the substance is in the attachment nobody kept
What good looks like without much effort
- Write down anything said on site that changes the works, the same day, and send it
- Date and file photographs as they are taken, not at the end of the job
- Record the reason alongside any assessment that differs from what was claimed
- Keep proof of service with each notice, not in a separate mailbox
- Use a system that records who changed what and when, so the history exists without anyone maintaining it
The last point is the one that scales. The first four depend on a person doing something conscientious during a busy week. The fifth happens whether or not anyone is having a good week, which is when the records that matter are usually being made.
The wider benefit
Most jobs never reach adjudication. The same records that would win one make the ordinary work easier: the surveyor covering a colleague's job, the client asking why a figure moved, the final account that reconciles without a reconstruction.
Which is a better reason to keep them than the one this article started with.
QScope records every change to every figure with the user and the timestamp, in a register no user can edit, so the contemporaneous record exists whether or not anyone thought to create it.