A global claim presents a total loss, sets out a list of events said to have caused it, and does not link the two. The claiming party asserts that the events together produced the loss and invites the reader to accept the whole.
It is a recognised approach and it is a weak one. The reason is structural: without the link between cause and effect, a single successful challenge to any part of the causation can undermine the entire claim rather than reducing it.
Why parties end up there
Rarely by choice. A global claim is usually assembled when the records will not support an itemised one, either because contemporaneous records were never made, or because so many events overlapped that separating them looks impossible.
Sometimes it reflects genuine complexity. On a heavily disrupted project the interaction between events can be real and difficult to disentangle. More often it reflects a project where nothing was recorded at the time and the analysis is being attempted from a programme after the event.
The distinction matters, because the first can sometimes be presented defensibly and the second usually cannot.
The vulnerability
The weakness in a global claim is that it stands or falls together. If any material part of the loss is shown to have a cause for which the claiming party is responsible, the whole basis of the claim is in difficulty, because the claim has not identified which part of the loss belongs to which event.
A defending party therefore does not need to answer every event. Finding one substantial contributor that belongs to the claimant, and showing the claim cannot separate it out, is often enough to put the whole thing in doubt.
That is a very poor risk profile for a claimant. An itemised claim of the same value, where one head fails, loses that head. A global claim of the same value, where one element fails, can lose everything.
What to do instead
Disaggregate as far as the records allow. Even partial itemisation is a substantial improvement, because the parts that can be linked stand on their own and only the remainder carries the global weakness.
Group by cause where individual linking is impossible. A cluster of events with a common effect on a defined area over a defined period is far more defensible than a single undifferentiated total.
Be explicit about what cannot be separated and why. A claim that says openly which parts are presented on a composite basis, and explains the reason, reads very differently from one that hopes nobody notices.
Records are the whole difference
A claim's structure is determined by the records that exist. Labour allocation by area and by week, possession dates, instruction dates, progress records and correspondence make itemisation possible. Their absence makes it impossible.
That is why records are a commercial issue rather than an administrative one, and why they cost almost nothing during the job and everything afterwards. A project that keeps them can present a claim in the strongest available form. One that does not is forced into the weakest.
This is particularly acute on phased and disrupted projects, where the volume of interacting events is highest, and it is set out further under complex phasing.
Defending one
The defending party's route is to identify the events for which the claimant is responsible and to demonstrate that the claim cannot allocate loss between them. That is analytical work rather than rhetoric, and it is done from the same records the claimant should have used.
It is also worth testing the loss itself. Global claims frequently price a total cost overrun as the loss, which assumes the tender was adequate and that nothing else contributed. Both assumptions can be examined.
Where a defence succeeds on those grounds, the outcome can be very one sided, which is why parties on both sides should understand the structure before committing to a position. The pattern is described in variation claim exposure.
The honest conversation
A claimant told early that their claim will have to be global has been told something important: the position is weaker than the value suggests, and settlement below the number is likely to be the commercial answer.
That conversation is easier in month one than in month twelve, and it is the point of testing entitlement and evidence before pricing anything, which is set out in testing entitlement before you argue about value.
Where the analysis genuinely cannot be done, saying so is more useful than producing a document that will not survive scrutiny.
What this means for you
If you are preparing a claim, disaggregate as far as you can and be explicit about the rest. Partial itemisation materially improves the position.
If you are defending one, look for the contributing cause that belongs to the claimant before you look at the quantum. It is usually the cheaper route to the answer. Our approach sits under final accounts and disputes.
Facing a claim with no causal detail?
Send what has been served. We will tell you what it actually establishes and where it can be tested.