Most claims that fail are not wrong. They are unproved, which from the far side of the table looks identical. Understanding what has to be shown, and in what order, is the difference between a position and a claim, and it is what a contractual entitlement review tests before anything is assembled.
The requirement is not forensic perfection. It is that somebody who was not there, reading what you produce without your explanation, can follow it and reach the same conclusion.
Three things have to be shown
That the event occurred, as a matter of fact, on dates that can be established from records rather than from recollection.
That the contract gives a route for that event, which is a question about the document rather than about the evidence.
And that the event caused the loss or delay claimed, which is the link that is most often missing and most often decisive.
Each is a separate exercise. Strong evidence on one does not compensate for weakness on another, which is why claims with impressive cost records sometimes fail completely.
Causation is the hard part
It is comparatively easy to show that something went wrong and that costs were higher than planned. It is much harder to show that the first caused the second.
Projects are disrupted by many things simultaneously, and a claim that attributes the whole of an overspend to one event invites the obvious question about the others.
The claims that survive are specific: this event, on these dates, had this effect on this activity, which cost this much. The ones that fail are general.
Where causation genuinely cannot be separated, that is worth knowing early, because it shapes what can realistically be claimed.
What counts as a record
Documents made at the time, for an ordinary project purpose, by people who were there. Site diaries, allocation sheets, measurement, programme updates, progress photographs, correspondence.
What counts for less: schedules prepared afterwards from invoices, narratives written for the claim, and witness recollection unsupported by anything contemporaneous.
The distinction is not about honesty. It is that a document created for the dispute is read as an argument, while one created at the time is read as evidence.
The records people wish they had kept
Daily allocation of labour and plant to activities rather than to the project as a whole. That single record converts most disruption claims from assertion into demonstration.
Programme updates issued and responded to, rather than maintained internally and never shared.
Instructions confirmed in writing when given, and notices served under the right provisions at the right time.
None of those is onerous. All of them are more likely on projects where somebody decided they mattered before anything went wrong.
Presenting it so it can be followed
Structure beats volume. A claim set out as a sequence of discrete issues, each with its facts, its contractual basis, its evidence and its value, is far more persuasive than a long narrative with appendices.
Cross reference to the documents rather than describing them. An assessor who has to search for the support will assume it is not there.
And separate what is proved from what is asserted. A claim that distinguishes the two is believed on the first part; one that presents everything with equal confidence is doubted throughout.
Naming your own gaps
Every claim has weak points. The question is whether you identify them or the other side does.
A claim that acknowledges a thin area and explains what can still be shown retains credibility. One that asserts uniformly loses it the moment a single assertion is tested and fails.
This is also a commercial tool. Knowing which parts are weak tells you where to concede in negotiation and where to hold.
What to do when the records are thin
Establish what does exist before concluding that nothing does. Records are frequently scattered rather than absent, and a systematic search produces more than people expect.
Look for documents created by others: the contract administrator’s records, the design team’s correspondence, the subcontractor’s returns. They are evidence too.
And be realistic about what can be proved. A claim scaled to the evidence is worth more than a larger one that collapses, because it survives examination and gets paid.
Proving it on different kinds of claim
A variation claim is usually the easiest to prove, because the instruction, the measurement and the valuation rule all produce documents.
A delay claim is harder, because it depends on the programme and on establishing what was actually critical rather than what was merely late.
A disruption claim is hardest of all, because the effect is spread across many activities and the evidence has to separate it from everything else happening at the same time. Each of the claims and dispute services is built around the kind of proof its subject actually requires.
Records somebody else made
On most projects the other side holds records that help you. Progress photographs, inspection records, minutes of meetings you attended, and correspondence sent to you all form part of the picture.
So do the subcontractors’ records, which frequently show the effect of an event more clearly than the main contractor’s own, because they are closer to the work.
Assembling those before concluding that the evidence is thin is worth doing properly. It is also considerably quicker than reconstructing the same facts from memory.
The cheapest version of all this
Records kept during the works, because somebody required them, in a form that makes sense to a reader. That costs very little at the time and it is the entire difference between a provable claim and an arguable one.
It is ordinary cost control and variations work rather than claims preparation, and it is done by people who do not expect to need it.
Where a claim is already being assembled from what exists, establishing what can actually be shown comes before deciding what to claim, not after.
Not sure whether the contract gives you a route?
A contractual entitlement review tests whether the contract gives a right to additional payment or time for what happened, before anybody argues about what it is worth.