In a previous article on claims under FIDIC contracts, we saw that form conditions the very existence of a claim: a perfectly well-founded entitlement notified out of time enters the turbulence zone of the time-bar. Let us assume that this first check has been done, that the deadlines have been verified and that nothing is time-barred. All that remains is… to write. And it is precisely there, facing the blank page, that the second act of the claim plays out.
Facing that blank page, four schools can generally be observed. The first is that of the inspired, who launch into lyrical flights and polish their rhetorical effects. The second is that of the so-called “legal” profiles, often shaped by a legal department or a law firm, who reproduce the codes of the formal notice and of pre-litigation. The third is that of the so-called “technical” profiles, who see everything through the prism of demonstration and often devote most of the letter to proving that they are right. The fourth, finally, is that of the score-settlers, for whom the letter becomes an outlet into which months of accumulated frustration are poured. Four very different styles, but one and the same underlying belief: a claim would be, above all, a writing exercise, and its quality would be measured by the pen.
Our practice leads us to the opposite conclusion, with a counter-intuitive observation: poor claim letters rarely lack style, they lack clarity. If writing quality long masked this reality, AI has levelled the field (upwards, in a sense) on spelling and grammar, which are now corrected in a matter of seconds. What remains to be done is what no tool will ever do in the contract manager’s place: the thinking and the interfacing.
A letter is first and foremost an act of communication
Before being a contractual document, a claim letter is a signal. There is a sender, a message and a receiver, and one certainty deserves to be stated from the outset: the signal will always be decoded. A letter sent is a letter read, interpreted, filed. The real question is therefore not whether the message will be received, but how it will be: as intended, or quite otherwise.
This reading grid is no mere image: it is, almost line for line, the model formalised by Claude Shannon and Warren Weaver in 1948 for telecommunications, which has since become the matrix of most communication theories. An information source produces a message, a transmitter encodes it into a signal, that signal travels through a channel exposed to a source of noise, and a receiver decodes it at destination. Transposed to contract management, the correspondence is striking: a project team as the information source, a contract manager as the transmitter, a letter as the channel, an addressee as the receiver. The model illustrates beautifully that one does not write a letter for oneself; one writes in order to be properly decoded.
It remains to be known by whom, since the receiver of a claim is in reality a plurality of receivers. Within the organisation receiving the letter, first: the primary addressee may well be the project manager, but the letter will circulate, will be escalated to a legal department, sometimes to general management, and each will read it through their own frame of reference. Over time, next: without this being either the objective or the wish, a claim letter may resurface months or years later in the hands of a DAB (dispute adjudication board), an expert or a judge, who will discover it with no project context whatsoever. A good letter therefore emits a signal that can be decoded by readers who share neither the same information nor the same time horizon.
Which leaves the last piece of the model, and probably the most useful one in practice: noise. Everything in a letter that does not serve the main objective (and, where relevant, the secondary ones) parasites it: non-factual adverbs, value judgements, stylistic effects, point-by-point replies that dilute the main argument among ten secondary ones. Reducing the noise means increasing the chances that the signal gets through as intended. The rest of this article is, at bottom, nothing but a declension of that principle, in three stages: capture, articulate, transmit.
Capturing information: a claim is built far from the keyboard
We will never repeat this often enough: the first skill for writing a letter or a claim is not an editorial one! A good claim begins with conversations: with planning, with the technical teams, with procurement, with the field. It is by multiplying exchanges with the right people that the contract manager collects the right information and, above all, becomes able to bridge elements that seemed unconnected. A late study, a delayed client decision, a scope change that went almost unnoticed: taken in isolation, these facts tell no story, whereas linked together they sometimes form the central argument of the case.
We thus occasionally come across contract managers who settle for asking the project manager for arguments to reply with, and who reuse them as they stand to build their letter. While this practice has the merit of being quick and effective for small letters with little at stake, it falls well short where a claim is concerned, since the project manager, however central, constitutes only one reading prism. It is precisely by crossing prisms and perspectives that the solid argument emerges, because each function holds a piece of the puzzle that the others cannot see.
This interface work is precisely the core of the contract manager’s trade. Whoever sits straight down at the keyboard deprives themselves of the claim’s raw material, and the stigmata show at first glance: the letter asserts but does not demonstrate, because the facts that would have allowed a demonstration were never collected. Conversely, a contract manager who has spent time at the interfaces arrives at the drafting stage with dated, sourced, cross-checked material, the very material that will later feed, on delay matters, structured analysis methods such as the Time Impact Analysis. Writing then becomes nothing more than shaping.
Articulating: the art of turning inputs into messages
Then comes the thinking work, a step often neglected by contract managers (for lack of time and/or of method). Indeed, the tedious collection work covered in the previous section produces inputs of all kinds: facts, opinions, quantifications, sometimes contradictory convictions. The frequent mistake consists in piling them into the letter in the order in which they arrived. The result is instantly recognisable: a letter you finish reading while wondering what its author wanted to obtain.
The question that must imperatively precede any drafting is therefore this one: what is the main message we want to convey? That question is best asked collectively, within the project team, before anyone picks up the pen. Agreeing on the main message, identifying the secondary messages, deciding what will not be said: this alignment work prevents the letters that go in every direction because everyone slipped their own topic into them.
Once alignment is secured, an individual task remains: sequencing. How to articulate these ideas so that the letter reads like a story, with a thread, a progression, a conclusion that flows from what precedes and, finally, a request (so often forgotten in letters)? This point matters all the more when replying to a letter received, for nothing obliges you to answer point by point a letter built in four points: some points deserve to be merged, others to remain unanswered, and the architecture of the reply belongs to whoever writes it. Answering within the adversary’s plan is already conceding them the framing of the debate.
Transmitting: a few elementary reminders
Then, and only then, comes the transmission itself, that is to say the drafting. This stage, well mastered by contract managers, nonetheless deserves three reminders.
The letter in service of the contractual relationship
The first is relational. A letter is not an isolated object: it is a vector of the contractual relationship, in the same way as a site meeting or a face-to-face with the client. It maintains that relationship, improves it or degrades it.
Yet we regularly observe situations that are hard to read for the other party: a cordial and constructive day-to-day project relationship, and threatening letters that seem written by somebody else. This gap is often explained by drafting delegated to a support function outside the project, which is not in the relationship and therefore cannot take it into account. These functions nonetheless have a precious role to play, only elsewhere: their distance gives them precisely the objectivity that those living the project lack, which makes them excellent reviewers, able to spot a weakness in the argument, to ask the questions that nobody in the team asks anymore, and to flush out the shortcuts that have become invisible from living the file too closely.
When a reviewer stops on a passage and says “here, I do not see what we are driving at”, there are only two hypotheses: either they are right on the substance, or the drafting took a shortcut by assuming information to be known when it is not. In both cases, the lesson is the same: if someone in-house, who knows the project a little, cannot decode the signal, the other party will not decode it any better. Drafting therefore belongs to whoever is in the relationship, while reviewing is the work of whoever has stepped out of it. This does not mean that raising the tone or shifting register is off limits when the situation demands it. It means that the tone of a letter is a decision, not a mood.
The facts, nothing but the facts
On substance, drafting a letter or a claim shares much with the legal approach: documenting the existence of a triggering event, establishing a causal link, relating it to the contract and presenting the consequences. One major difference remains, however, as to form: where a legal pen may, rightly or wrongly, choose to add an emotional charge to its point, the contract management letter gains from remaining exclusively factual.
Intensity adverbs and emotional formulas add no information and weaken the point. Here are a few examples that every one of you will have come across in your contract management practice: “after numerous reminders”: numerous, meaning what? Two, four or six? Or the famous “we are deeply disappointed by the content of your letter”: disappointment is not a contractual fact. The same vigilance applies to value judgements: “your performance is poor” has no place in a letter. A service is compliant or non-compliant, a defect is minor or major, and so on.
To illustrate with a before-and-after, here is first a “noisy” version to avoid: “Despite our very numerous reminders having gone unanswered, we regret to note that your performance remains unacceptable.” And its factual, far more effective counterpart: “Notwithstanding our reminders of 3, 17 and 28 March (ref. C-041, C-052 and C-067), the major defects notified on 15 February remain uncleared to date, contrary to Article 9.3 of the contract.”
Trace, document, follow up
Finally, let us close with a point as trivial as it is essential, what we call “documentary hygiene”. Every letter should carry a date, a proper reference, a recall of the previous correspondence dealing with the same subject: this chaining seems tedious; it is in reality strategic.
It is what allows the story to be reconstructed effortlessly, months or years later. We find here the future reader mentioned above: a well-chained correspondence file reads like a coherent narrative, where orphan letters force everyone to rebuild the context, with all the risks of interpretation that this entails.
The contract manager, a strategic interface rather than a pen
At the end of this article, one conclusion stands out (and that was our objective): in a good claim letter, the writing is the last step, and probably the least discriminating one. What separates a letter that obtains a result from a letter that feeds the conflict plays out upstream, in the quality of the collection, in the clarity of the message, in the consistency with the relationship.
AI, by making correction and formatting accessible to all, has not devalued the contract manager’s work on claims. Quite the contrary: it brings us back to our true trade, to our raison d’être, that of a strategic interface that connects information nobody else connects, and turns that material into a signal that its various receivers, those of today as well as those of tomorrow, will know how to decode.
