Project records are useful during delivery as well as in a later dispute. Their value lies in showing what happened, what the team knew and why work or decisions changed—not simply in the number of photographs or forms retained.
Record enough to explain the work
A project can produce thousands of photographs, emails and diary entries without being able to explain why an activity took longer than planned. The problem is often the missing connection between the event, the work affected and the response on site.
Records should help the team manage the project now and explain its decisions later. They need to be reliable, easy to find and proportionate to the work. Adding more fields to a form will not help if the important information is still missing.
Make the daily record specific
“Drainage delayed” provides little assistance. A useful entry identifies the workfront, the planned activity, what prevented it proceeding, the period affected and what the team did instead. It distinguishes an observed fact from an explanation that still needs checking.
For example, a record might state that the crew assigned to drainage at chainage 420 could not start because the service location had not been confirmed. It should also record whether the crew moved to another workfront, which plant remained idle and when the outstanding information was received.
That level of detail allows someone to check the account against drawings, instructions, timesheets and the programme. It does not decide contractual responsibility. It gives the people making that assessment something dependable to work with.
Connect resources to activities
A timesheet showing eight people on site establishes attendance. It may not explain what those people did. Where an issue is likely to affect cost or productivity, record the relevant activity, location, hours and output alongside the resource allocation.
Choose cost codes that the site team can use consistently. If every additional activity is charged to a general variations code, separating individual events later will be difficult. A short event reference used across site records, purchase orders and the change register is often enough.
Do not expect a photograph to fill every gap. Record its date, location and subject, and retain the original file. A close-up of an obstruction is more useful when another image or sketch shows where it sits in relation to the work.
Keep the programme history
Retain the programme versions used to make decisions, together with the information supporting their updates. Actual start and finish dates, progress achieved and changes in sequence should be traceable. Overwriting the previous file removes part of the project’s history.
When a sequence changes, record why. A revised programme may show that one activity moved behind another, but it may not explain whether that resulted from an instruction, unavailable access, a resource decision or a recovery measure.
The explanation should be made when the decision is taken. A short note agreed with the planner and supervisor is preferable to trying to reconstruct the reasoning several months later.
Record decisions and outstanding actions
Meeting minutes should distinguish a discussion from a decision. Identify what was agreed, who had authority to agree it and which actions remain outstanding. If the contract requires a separate notice or instruction, issue it through that procedure rather than assuming the minutes will suffice.
Where accounts differ, record the disagreement without rewriting the underlying event. If an entry needs correction, preserve the original and add a dated explanation. An apparently tidy record that conceals how it developed is less useful than a clear audit trail.
Keep factual project records separate from privileged legal advice and follow the organisation’s controls for personal, confidential and commercially sensitive information. Access to the record should suit the purpose for which it is held.
Test whether the system works
Take one current issue and ask someone who was not involved to follow it through the records. Can they identify the instruction, affected work, resources, programme position and current assessment? Can they distinguish agreed matters from assumptions?
If they cannot, fix the missing connection while the information is still available. This is a more useful test than checking whether every document has been uploaded. A filing system can be complete and the project history still be unclear.
Allocate responsibility for the records that matter, including when staff leave or subcontractors finish. The aim is a dependable account of the work that the team can use without relying on one person’s memory.
