How to Structure a Proposal So the Part That Matters Is Read First

The standard proposal order is background, understanding of the brief, approach, methodology, team, timeline, pricing, terms. It is close to exactly backwards, and it is backwards for a reason worth understanding: it is ordered by the sequence in which you did the work, not by the sequence in which anybody needs to read it.

Order by what the reader has to decide, not by what you had to do

Journalism solved this a century and a half ago and called it the inverted pyramid: the conclusion first, then the detail that supports it, then the background, on the assumption that the reader stops at an unpredictable point and every reader who stops should still have the most important thing.

A proposal has exactly that shape of readership. Some people read all of it. Most read the top. One person reads only what somebody else pasted into a message. Ordering for the reader who finishes is optimising for the minority.

So, in order:

1. What you will do. One sentence. Their words for the problem, not your words for your service.

2. What it costs. The number, or the range and what moves it.

3. When it is done. A date, or a duration from a start.

4. Why you rather than the other quote on their desk. Two or three lines, specific, checkable, and not adjectives.

5. What happens next. The smallest action that moves this forward, named and easy.

6. Everything else. Scope in detail, assumptions, exclusions, method, team, case studies, terms. Long is fine here. Nobody is blocked by it.

Items one to five are the argument and should fit in the space a person will read without deciding to. Item six is the evidence and can be as long as it honestly needs to be.

The three sections that are read least

These are the ones that feel obligatory, take the longest to write, and sit exactly where the reader is deciding whether to keep going.

• “Our understanding of your requirements.” You are restating their brief to them. They wrote it. What this section is actually for is proving you listened — which one specific, slightly unexpected observation does better than a page of paraphrase.

• The company history. Founded in, offices in, values are. It is the section most likely to be skipped and most likely to be defended internally. Move it behind the argument and see whether anyone notices.

• The methodology diagram. Discover → Define → Deliver. Every competing proposal on the desk has one and they are all the same, which means it cannot differentiate you and is costing you a screen.

None of these has to be deleted. All three have to stop being between the reader and the decision.

Write the top for the person you did not send it to

Your document will very often be read by somebody who never received your email. It gets pasted into an internal thread, and from that point it is being read by a person with no covering message, no prior conversation and less patience.

That reader is frequently the one who decides. Everything above is really a way of writing for them: the top of the document has to work with the email stripped off it. What a forwarded link can and cannot tell you is what that journey looks like from your side of it.

How to test an order, given that you cannot see sections

This is where honesty about the tooling matters, including ours. Quixli records that a document was opened, by whom where that can be established, for how long, and how many times. It does not record which section anybody read. There is no scroll map and no per-heading heatmap, and any claim that reordering your document “increased time on section 3” is not something the data can support.

So you test structure at the document level, across sends, like this:

1. Change one thing. Move pricing forward, or cut the methodology section. One change, not a redesign, or you will not know which part did it.

2. Send ten comparable proposals with the new order. Same kind of buyer, same kind of deal, or the comparison is meaningless.

3. Compare three things, not one. Median first-session reading time, the share of documents that get a second visit, and the share that get any reply at all. Reading time alone can fall for a good reason — a shorter document read completely — so it is the least useful of the three on its own.

4. Expect the return visit to move first. If the decision is easy to find, the person who comes back to check a number finds it and stays; if it is buried, they give up. Return visits are the most sensitive of the three measures to ordering.

The confound to keep in mind: your second ten proposals went to different people in a different month. You are not running a controlled experiment and you should not report it as one. What you have is a direction, which is all you needed. What a second open actually means covers reading that return-visit signal properly.

One structural rule that survives every situation

Nothing that the reader has to have should be below something they might not want.

That single rule generates most of the advice above without needing any of it. It is also the rule that tells you when to break the convention in the other direction — a technical audience that will not accept a number before a method genuinely does want the method first, and you should give it to them.

The two decisions that follow from ordering are what goes on the very first screen and where the price sits. And what Quixli can and cannot show you about a document is the limit the test above is written inside.