How to Write a Proposal That Wins the Project

In the consultant Alan Weiss’s view, the sale happens before the proposal is ever written. He describes a proposal as a written statement of the agreement already reached in earlier conversations with the buyer, and that is the most useful way to think about how to write a proposal.¹ Hold a discovery call first, then put what you heard into eight sections, in this order:

1. Restate the client’s problem in their words.
2. Describe the outcome they will get.
3. Outline your approach.
4. Set the scope and say what is excluded.
5. Give the timeline.
6. Price it with two or three options.
7. Add proof that fits this client.
8. End with the next step and an expiry date.

Send it within a day or two of the call, and decide how you will follow up before you press send. The same order works for a business proposal to a company and a project proposal to a private client. Anyone still deciding whether starting a business suits them should count proposals among the routine tasks that come with it.

Why the discovery call comes before the proposal

A discovery call gives you the client’s problem, goals, budget range, deadline and decision-makers in their own words, and a proposal without those is a guess. Australia’s government business site tells businesses to know what the customer’s goals are before quoting, and the UK Government Commercial Agency warns suppliers against generic, recycled responses.²³

A client reads your proposal against what they told you. When the problem, the outcome and the price range match the conversation, reading it feels like recognition, and saying yes feels like keeping a decision they have already half made. When it does not match, every section invites a new round of questions.

A hotel owner who emails “Can you send us a proposal for marketing?” has given you almost nothing to write. A better reply offers a 30-minute call and five questions: What prompted this now? What happens if nothing changes? What would count as success six months from now? Who else has a say? What range have you set aside, and when do you want to start?

It is a written statement of the conceptual agreement already gained in prior discussions with the economic buyer.

Treat a request for a proposal before any conversation as a request for a call. If you find yourself writing to someone you have not spoken to, the proposal is too early.

How to write a proposal in eight parts

Each part answers a question the client will ask whether you address it or not. The running example is made up: a freelance marketer writing to a small family-run hotel whose direct bookings have fallen.

1. Problem: what the client told you, in their own words
2. Outcome: what will be different, and how you will both measure it
3. Approach: the few main steps, in plain language
4. Scope: what is included, and what is not
5. Timeline: dates, and what you need from the client by when
6. Price options: two or three, built around what the client asked for
7. Proof: one or two specific, checkable examples that match this client
8. Next step and expiry: how to say yes, and until when the offer holds

1. Restate the client’s problem in their words

Open with two or three sentences that repeat what the client said, using their phrases. “You told us direct bookings have dropped over the past year and that commission to booking sites now takes too large a share of each stay.” This shows you listened, and it gives anyone who missed the call the reason the project exists.

2. Describe the outcome they will get

The outcome says what will be different and how you will both measure it. The UK agency’s bid guide asks suppliers for measurable outcomes and buyer-specific benefits.³ For the hotel: “A larger share of rooms booked through your own website, tracked monthly from your booking system.” Promise the measure and the work; a promised number is a guess about things you do not control.

3. Outline your approach

Three to five plain steps are enough: audit the booking page, fix the slowest steps, then run a three-month email campaign to past guests. Clients buy the result. Detail about tools and methods can wait for the kickoff.

4. Set the scope and say what is excluded

List what is included and, just as clearly, what is not: here, photography, a new website and paid ads. Writing a scope of work that keeps a project from sprawling is its own skill; in the proposal, a short version is enough, and the full scope can carry over into the contract.

5. Give the timeline

Give dates counted from signing and tie each one to what you need from the client. “Audit in week 2, provided we have access to your booking system by the end of week 1.”

6. Price it with two or three options

Build the middle option around what the client asked for, a smaller option that does less, and a larger one that does more. “Booking-page fixes only”, “Fixes plus three months of email”, and “Fixes, email and a quarterly review”. Each option should change what the client gets. A smaller option that only cuts the price undercuts the middle one. Whether each price comes from hours or from pricing on the value the work creates is a separate decision; the options are about presentation.

7. Add proof that fits this client

Choose one or two examples that match the client’s situation over a page of logos. The UK agency’s advice to bidders is to be specific, with relevant examples and data behind every claim.³ Name the kind of client, the work you did and the result, and only use names and figures you have permission to share.

8. End with the next step and an expiry date

Say exactly how to go ahead: choose an option, sign, pay the deposit, and the start date. business.gov.au lists a quote expiry date and space for the customer’s acceptance among the things a written quote should include.² An expiry date protects your calendar and your prices, and it gives your follow-up a natural deadline.

Read the draft once more against the call notes, then tick through the eight sections.

Before you send the proposal

• A discovery call happened, and the problem is in the client's words
• The outcome names how it will be measured
• The approach fits in three to five steps
• Exclusions are listed, not just inclusions
• Dates are tied to what the client must supply
• Two or three options, each changing what the client gets
• Proof matches this client and is cleared to share
• One clear next step and an expiry date

Two or three price options work better than one

Offering a client a choice between two or three versions of the work, instead of a single fixed price, fits what consumer research has found about how people choose. In a 2013 series of lab studies, Daniel Mochon found that a lone option makes people want to keep searching, so a product can be chosen more often when an alternative sits beside it.⁴

The mechanism matters for freelancers. A single price invites the question “What else is out there?”, which sends the client to find other quotes. Two or three options turn it into “Which of these?”, a comparison you control.

People also lean toward the middle of a range. A meta-analysis of 142 experimental observations found middle options were picked significantly more often than the extremes, although the pull was weaker in studies built on price-quality trade-offs, the kind of trade-off a tiered proposal asks for.⁵ More choice is not always better either: a meta-analysis of choice overloadchoice overload: The idea that being offered more options makes choosing harder, leaving people less likely to choose or less satisfied with what they pick. Meta-analyses find little effect on average, with overload appearing only under particular conditions, such as complex options or a hard decision.Full entry in the glossary found that large assortments are more likely to backfire when the options are complex and the decision is hard.⁶

Myth

The more options you offer, the easier it is for a client to find one they like.

Fact

A lone option makes people keep searching, but large, complex sets make choosing harder. A small, clear set sits between the two.

All of this comes from consumer experiments, mostly about products, and none tested service proposals. Use it as a reason to offer a small, clear set, and make the option you recommend the one the client asked for.

Your first number sets the range

The price in a proposal is a first offer, and research on negotiation shows first offers act as anchorsanchoring: The tendency for an estimate or decision to be pulled toward a number seen beforehand, even an irrelevant one. It is one of the effects that held up when large teams ran the original experiments again.Full entry in the glossary. A 2025 meta-analysis by H. M. Petrowsky, D. D. Loschelder and 13 colleagues found that first offers pulled final agreements toward them, while ambitious openers paid for it in lost deals.⁷

The study

Moderate evidence

Across 90 negotiation studies, the first number pulled the deal toward it, and ambitious openers cost deals

A preregistered meta-analysis of 374 effects from 90 studies, with more than 16,000 participants, found a general advantage for the side that makes the first offer and a strong link between the size of that offer and the final agreement. It also found that ambitious first offers led to more impasses and made the people receiving them feel worse about the deal. Two further preregistered experiments repeated both patterns, and the effects shrank as negotiations covered more issues.⁷

The studies are negotiations, not written proposals; this article relies on the paper’s summary because the full text could not be fetched, so details such as the countries sampled are not reported here. The finding on complexity is relevant: a proposal with scope, dates and several options is a multi-issue deal, where the anchor’s pull is weaker.

Your number frames the conversation, so do not open low out of nerves. An inflated price is a different risk. The same meta-analysis found that ambitious openers lose more deals and leave the client feeling worse, which matters when you want repeat work. Price each option at a figure you can explain from the outcome.

Specific claims give the client something to check

Specific wording gives a client something to verify, which is reason enough to use it, although the evidence that concrete language alone makes claims seem truer is mixed. In a 2010 set of experiments, Jochim Hansen and Michaela Wänke found that statements written in concrete language were judged more likely to be true than the same content written abstractly.⁸

A 2019 preregistered replication with several hundred participants found smaller effects than the original in both of its experiments, neither clearly different from zero, and its authors concluded the overall pattern was inconsistent with the original study.⁹ So the persuasive boost from concrete wording may be small. The plainer benefit holds anyway: a specific claim can be checked, and a vague one cannot.

The UK agency’s bid guide warns against generic statements such as “customer service is at the heart of what we do” and asks suppliers to show how instead.³ “We have years of hospitality experience” tells the hotel owner nothing they can check. A sentence such as “We rebuilt the booking page for a 20-room guesthouse and tracked direct bookings for six months”, if true, tells them what to ask about.

How long should a proposal be?

A proposal should be long enough to cover the eight sections and no longer; for a small project, that can be two or three pages. Weiss says his own proposals typically run to two and a half pages, with no legal boilerplate and no staff biographies unless the client requires them.¹

The only large data set on length belongs to a proposal-software vendor. Proposify’s 2025 report, covering proposals sent through its platform in 2024, found that winning proposals averaged 11 pages and losing ones 13.¹⁰ Those figures carry three caveats. The sample is the company’s own customers; the comparison is a correlation; and the report has not been peer-reviewed. Proposify itself says the right length depends on the deal’s size and complexity, so longer proposals may lose because they go to bigger, harder deals.

Both sources point the same way. For the hotel, a page on the marketer’s history, a list of every tool they use and a long terms section can all go; the eight parts can fit on three pages. Cut anything the client did not ask about and that does not help them decide.

How do you follow up on a proposal?

Follow up on a schedule you set before sending: agree on the call when the proposal will arrive and when you will talk it through, then follow up on those dates. No independent study has tested follow-up timing for proposals, so the routine below is practitioner advice.

Proposify’s report offers two hints from its platform data. Deals that closed took about two days on average from the first time the client opened the proposal, and close rates were twice as high when more than one person at the client viewed it.¹⁰ Both figures describe its users and show association, not cause. They still suggest two habits: send while the call is fresh, and ask who else needs to read the proposal.

1. 1Discovery callagree when the proposal will arrive
2. 2Send in one or two daysoffer a short walk-through call
3. 3Check in after two or three daysask what questions have come up
4. 4After a weekadd something useful, such as an answer to a question they raised
5. 5Before the expiry dateone last note, then close the file

When the client asks for changes, treat it as a good sign; Proposify’s report says proposals that get revised have a higher close rate on its platform.¹⁰ When the expiry date passes without a reply, send one short note, then move on.

How much weight the evidence can bear

The bottom line

Most of the work of winning a project happens on the discovery call, and the proposal puts what you agreed there on paper. Before your next proposal, book the discovery call, take notes in the client’s words, and send the eight sections within a day or two with an expiry date and a time to talk it through.