Open a report with what you are recommending, then give the reasoning, then put the detail somewhere it can be skipped. A senior reader gives a document a short look before deciding whether to continue, and a recommendation on page four has been placed behind that decision.
This is the exact inverse of the essay shape, which builds through the evidence and arrives at the conclusion at the end. Most people who write well at work were trained in the essay shape, at school and at university, and were never told that the professional version reverses it.
Eight rules for writing that gets a reply
- put the ask in the first two lines. English business writing front loads the request, and burying it under context reads as evasive rather than polite.
- use the subject line as the summary, so a reader who opens nothing else still knows what is wanted.
- one email, one subject. Two requests in one message reliably get one answer.
- Please could you confirm by Thursday. A named deadline is not rude and its absence is the commonest reason nothing happens.
- bullet the options when asking somebody to choose, which makes replying a one word job.
- short paragraphs, plenty of white space. A dense block is read as a wall and postponed.
- As discussed, referring back to a conversation, which anchors the message and removes the need for a paragraph of context.
- Apologies for the delay, which is the standard opening for a late reply and is preferred over an explanation nobody asked for.
Two shapes, and the training points the wrong way
An academic essay withholds the conclusion because the argument is the product: the reader is there to follow the reasoning, and arriving at the position too early looks like assertion rather than analysis.
A report is read by somebody deciding whether they need to read it. The product is the decision, not the reasoning, and a reader who agrees with the recommendation in the first line may never need the reasoning at all. Putting it first is not a shortcut for lazy readers; it is what makes the document usable by the people it was written for.
Both shapes are correct in their own place, which is why this is hard to unlearn. Nothing about the essay shape is wrong. It is simply aimed at a reader who has already committed to reading, and at work nobody has.
Three layers, and each one is complete
The useful structure has each layer standing alone. A first paragraph that could be the whole document. A body that gives the reasoning to somebody who wants it. An appendix holding the working, the data and the alternatives considered, for the one reader in ten who will check.
Written that way, a document survives being read at three depths, which is what actually happens to it: one person reads the first line, one reads the body, and one goes through the appendix looking for the assumption you made about next year's costs.
Where it stops: the well-built argument nobody reaches
The stall is writing the good version of the wrong shape. The document is well evidenced, carefully sequenced and genuinely persuasive from beginning to end, and the decision was made before page two.
What comes back is not criticism of the writing. It is silence, or a meeting where somebody asks a question the document answered on page five, or a decision that went the other way with no explanation. None of that names the structure, so the writer's next document is more thorough, and thoroughness is the thing that was already excessive.
The tell is length. If the important sentence in your document is more than a screen from the top, the shape is wrong regardless of how good the sentence is.
Guidance written for people who are skimming
The GOV.UK content design guidance is free, public, and unusually good on exactly this, because it was written for an audience that has no obligation to keep reading and often no choice about being there. It is specific about front-loading, about what belongs in a summary, and about the words that make a sentence longer without making it clearer.
It replaced an internal style guide, which every large organisation has and almost nobody reads, because it was written to settle arguments rather than to be useful.
Its limitation is the reading level it targets. Public-service writing aims deliberately low, and a technical proposal to specialists does not need to, so take the structure and the front-loading and do not take the vocabulary ceiling.
What it looks like when it goes wrong
The recognisable failure is the proposal that gets a reply asking what you are actually proposing. The document said so, at length, in the middle.
The second is a summary that summarises rather than recommends. This report examines three options for the migration is a description of the document; the second option is the one to take, for two reasons is a first line. The first version is what most executive summaries contain and it is why most of them are skipped.
The claim-first principle at the level of a single paragraph, and the test that finds paragraphs without one, is one idea, an opening line, and how the next one connects, and the same front-loading applied to a message somebody has to act on today is openings, closings and the register each situation asks for. Both sit under the English that decides whether you get the role.
What a document cannot do is answer back, so the reasoning that was skipped in the reading gets demanded out loud instead, usually by the one person who did read the appendix: signposting, transitions and handling the questions after.