Delay claims are decided on what the records show about what was holding the work up, week by week. Those records are cheap to keep and almost never kept properly, which is why so much money is spent reconstructing them afterwards during extension of time claim preparation.
Nothing in this list is sophisticated. The difficulty is doing it every day on a site where there is always something more urgent.
The site diary
A daily entry recording what was worked on, what was not, and why. The second half is the part that matters and the part usually left out.
Entries written the same day carry weight. Entries written up at the end of the month do not, and an experienced reader can tell the difference.
It is the cheapest and most persuasive record available on any project, and it takes ten minutes.
Resource allocated to activities
Labour and plant recorded against activities rather than against the project, weekly at minimum.
This is what allows productivity to be shown and what makes a disruption claim possible at all.
It is also useful management information, which is the argument for keeping it on projects where no dispute is expected.
Progress against the programme
Actual start and finish dates for activities, recorded as they happen rather than estimated later.
Percentage complete assessments are too coarse to support a delay analysis and will be challenged as such.
A marked up programme each month, dated and kept, is worth more than a narrative report of the same length.
Information and instruction records
What was requested, when, what was received and when. The gap between the two is frequently the whole claim.
A simple register maintained by the project team serves this purpose and removes the need to reconstruct it from correspondence.
Where the programme shows information requirements with dates, the register writes itself against them.
Records of what stopped
An activity that could not proceed, with the reason and the duration, recorded at the time.
This is the record that answers concurrency arguments, and its absence is why those arguments succeed so often.
One line in the diary is enough, provided it says what was prevented and by what.
Photographs with dates
Useful for conditions: access blocked, preceding work incomplete, areas congested, weather.
Not useful for productivity, and a submission that relies on them for that invites the obvious answer.
Filed by date and location rather than in one folder, otherwise they are unusable two years later.
Subcontractor records
Frequently better than the main contractor’s, because the subcontractor was measuring its own gangs against its own prices.
Worth asking for while the relationship is live rather than after a dispute has started.
Where a subcontract delay claim arrives, the records behind it may support the main claim as well.
What not to bother with
Long narrative monthly reports that describe the project without recording anything specific enough to be used.
Correspondence written for the file rather than to achieve something, which reads exactly as it was written.
Claim diaries kept separately from the project records, which look prepared and are given correspondingly little weight.
Who should keep them
The site team, as part of the daily routine, with the commercial team checking weekly that it is happening.
Records kept by the commercial team alone are written at a distance from the work and show it.
Where the site team sees the records used in a monthly conversation, they get kept. Where they disappear into a file, they stop.
Making it survive the project
A single structure, agreed at the start, so records are where the next person expects to find them.
Copies held somewhere other than a personal email account, because people leave and accounts close.
This is dull and it is the difference between a claim prepared in three weeks and one prepared in three months.
Starting late
On a project already in difficulty, the records can start today and the period from today is protected even if the earlier period is not.
That is worth doing the moment a dispute looks likely rather than when it arrives.
Reconstructing the earlier period from what exists is a separate exercise and should be identified as a reconstruction.
What it costs
Ten minutes a day on site and an hour a week in the commercial office, which on any project of size is immaterial.
Against that, a claim that can be prepared from records rather than from argument, and a defence that can be answered.
It is the clearest example in commercial management of a small habit with a disproportionate return.
The conversation worth having at the start
Agree with the site team what will be recorded, by whom, and where it will be kept, before the project is under pressure rather than after the first difficulty.
Show them once what the records are for, with an example of a claim that turned on a diary entry, because people keep records they understand the purpose of.
Then look at them monthly in a meeting somebody attends. Records that are never read stop being kept, and the month they stop is usually the month they were about to matter.
Where it fits
Records are not claims work. They are what makes claims work possible and short.
Where a claim has to be built from thin records, that is initial dispute review territory and the first question is what exists.
Keeping them during the works belongs with cost control and variations, which is where the return on them is actually earned.
Everything else in this field is an attempt to recover from not having done it, and none of those attempts is as cheap.
Running late with nobody agreeing why?
Extension of time claim preparation sets out the events relied on, their effect on the programme and the contractual route to relief, in a form that can be assessed rather than argued over.