For 33 years we have delivered technical reports, and there is one lesson that repeats itself more regularly than any other: a report that cannot be decided upon is not a finished report. The calculations may have been done correctly, the right standards may have been consulted, the methodology may have been followed rigorously. If it is not clear at the end of the document what needs to be done on Monday, the technical work has not been completed — it has been interrupted.
This is a problem that is not fixed in the writing. Almost never. It is fixed in the analysis.
For whom is a technical report really written
An environmental or energy engineering report is not written to demonstrate that we know. It is written so that someone can decide. And who reads it often is not an engineer: it is a management committee that must approve an investment, a technical office of the administration that must resolve a file, a financial entity that must assess a risk, a law firm that must defend a position before an environmental body.
They have one thing in common: little time and a lot of responsibility. The report that arrives on their desk will compete with twenty others for the same weekend, and it will be judged by the clarity with which it helps them to act, not by the elegance with which it presents the method.
This has technical consequences that affect how the document is structured from the first page.
Three rules that we apply to each report
1. The conclusion comes first, stated in terms of decision. What needs to be done, when, and what happens if it is not done. This is the first page. The complete and replicable calculation goes in the annexes, so that anyone who wants to verify it can easily find it, but it is not imposed on those who have to make the decision.
The reason is not aesthetic, it is functional: a decision-maker who has to go through 30 pages of methodological development before reaching “what we need to do” either does not get there, or arrives tired and biased. In either case, the report does not do the job for which it has been paid.
2. Uncertainty is written down, not hidden. A figure without a confidence range or explicit hypotheses cannot be defended six months later when someone — a verifier, a competitor, an administration, the client themselves — looks at the document again. And a figure that cannot be defended is a figure that cannot be used to make a lasting decision.
This means that each number in the report carries alongside it the conditions that make it verifiable: what basis, what period, what hypotheses, what range. Not to show that we know how to do it, but so that the number survives its own delivery.
3. The discarded alternatives are noted. A report that presents a recommendation without showing the options that were discarded leaves open the possibility that someone might propose them again in two years. Knowing why an alternative is not worthwhile — with the number that justifies it — is often more useful than the main recommendation itself.
The discarded alternatives are not an academic appendix: they are an active defence against the repetition of past mistakes.
The quality criterion: the Monday test
There is a simple check that we always apply before delivery: if after reading the report the client does not know what to do on Monday, there is something unresolved. And it is almost never in the wording. It is in the analysis.
This means that the final review of a report is not a style review. It is an operability review. If we read it as if we were the recipient and do not know how to act tomorrow, we need to return to the analysis — there is no need to rewrite the conclusion to make it “clearer”. Rewriting a conclusion without substantial clarity only disguises the problem.
Why this is Rigor, not stylistics
When we say Rigor as an AUMA value, we do not refer to a statement of intent, but to a discipline: that a figure or our criterion can be questioned by a third party and withstand. An external verifier, an administration, the client's technical department, a competitor interested in finding loopholes. If the report does not allow reconstructing how the number was reached — with unit, boundary, period, and explicit hypotheses —, the Rigor is not there. What exists is an aesthetic finish.
Thirty-three years without a matrix and without any product to place have left us with only one thing to defend: that what we sign can be verified. This directly ties in with methodological honesty: we prefer a figure with all its conditions written down rather than a round figure. Because the round one is memorable, but not decidable.
Methodological transparency is not about slowing down, it is about enabling
A frequent objection when someone sees our reports for the first time is that they incorporate “too many conditions”. It would translate, in a corridor conversation, as “don’t complicate yourselves, just put the figure”. The problem is that the figure alone, without conditions, does not survive the first audit.
Methodological transparency does not slow down a decision. It is what allows the decision to be made — and, more importantly, that it can be made with the assurance that it will withstand the next requirement, the next bank, the next manager who looks at it.
What we do
A completed report is a report that decides: it has a conclusion at the front, it has the conditions next to each figure, and it has the discarded alternatives available. Not because it looks nicer, but because this is what makes what we sign defensible before those who challenge it.
If you have a project where the decision is blocked due to a lack of criteria — not due to a lack of analysis, but due to a lack of decidable criteria — let’s talk about it. A two-hour session is often enough to see what is missing from the report so that your team can act next week.