Most people write the way they thought. Here is the background, here is what we examined, here are the considerations, and therefore here is the recommendation.
That structure mirrors the reasoning process and it fails with the specific audience it is usually aimed at, which is somebody with authority and eleven minutes.
Why the build fails
A reader who does not yet know where a document is going cannot evaluate anything in it. They read the background without knowing which parts will matter, the analysis without knowing what it is establishing, and they hold all of it in suspension waiting for a resolution.
This is expensive, and busy readers respond to expense by skipping. They jump to the end, find the recommendation without the reasoning, and form a view on the least supported version of your argument. Or they stop after two paragraphs and ask you what it says, which means the document did not work and you are now delivering it verbally after all.
The build also carries an implication nobody intends: that the reader must be walked to the conclusion rather than trusted to assess it. Senior readers notice this and find it wearying.
The order that works
Recommendation. Then the two or three reasons it is right. Then what it costs or risks. Then the detail, for whoever needs it.
We should replace the Nampa roof this fiscal year rather than deferring to FY28. Three reasons follow, then the objection, then the numbers.
A reader now knows within one sentence what they are being asked to decide, and every subsequent paragraph is being evaluated against a known claim. They can stop whenever they have enough, which for a decision they trust you on might be the first line, and for one they are sceptical about might be the appendix.
The structure serves both readers without requiring you to know in advance which one you have.
Put the objection in yourself
The instinct is to omit the strongest counterargument or bury it. This is a mistake on every dimension including the self-interested one.
A reader who finds the weakness you did not mention concludes either that you missed it or that you concealed it. Both are worse for you than the weakness itself. A reader who finds it stated plainly in your own document concludes that you looked hard at this, and the rest of the analysis gains credibility rather than losing it.
The form is one short paragraph. The case against this is that the levy vote is in November and the expenditure will be visible. I think the leak risk outweighs it, and here is why. That is stronger than any version omitting the first sentence.
Where the build is correct
Conclusion-first is a default, not a rule, and there are situations where the reverse genuinely works better.
When the conclusion will be rejected on sight by somebody who needs to be walked through the evidence before they will consider it. When the reasoning is the actual product, as in an analysis whose value is the method. When you are writing to think rather than to transmit, which is legitimate and should not be sent.
And in genuinely bad news to a person, where opening with the conclusion is correct but opening with the reasoning behind it is not. That is a matter of a sentence or two of framing, not of restructuring the document.
The subject line is part of it
Most professional writing now arrives as email, and the same principle applies one level up.
Nampa roof is a topic. Recommend replacing Nampa roof this FY, decision needed by the 14th is a message. The second tells a reader what the email is for, whether it requires them, and when, before they open it.
Any email that asks for something should say so in the first line, not the fourth paragraph. The request buried under context is the single most common reason things do not get done, and the sender invariably experiences it as the recipient not reading carefully.
The drafting order is different
None of this means you should think conclusion-first. You cannot; the conclusion is the output.
Write the analysis first, in whatever order it comes. Then find the sentence that states what you concluded, move it to the top, and cut the material that was scaffolding for your own reasoning rather than evidence for the reader’s.
That last step usually removes a third of the document. What is being removed is the record of how you got there, which felt essential while writing and is of no interest to anybody else.
Edited by Patrick J. Wolf, PhD