How to Write a Marketing Proposal That Wins Clients

A marketing proposal is not a document about your agency. It is a short, specific argument that a particular business has a particular problem, that the problem is worth fixing, and that you are the safest people to fix it. Most proposals fail because they get that order wrong. They open with the agency's history, move to a list of services, and arrive at a price before the reader has any reason to care about it.

This guide shows you a better way to write one. It covers what to include, how long a proposal should be, how to present pricing without scaring people off, what to do once it has been sent, and where most agencies quietly lose deals they should have won. Where we quote numbers, we link to the source, and we say so plainly when a figure comes from a company with something to sell.

The best proposals feel like a continuation of a good conversation, not the start of a new one.

What a marketing proposal actually needs to do

A proposal has four jobs, and only four. It must show that you understand the business. It must make the cost of doing nothing feel real. It must present a clear plan that a busy owner can follow. And it must make saying yes easy.

Everything else is decoration. A proposal can be beautifully designed and still fail all four jobs. An ugly one that does all four will often win, though you should not have to choose.

It helps to think about who is really reading. Proposals are rarely read once by one person. An owner skims it on a phone between jobs. A partner or manager reads it that evening. Someone asks, "What are we actually getting for this?" Your document has to answer that question when you are not in the room.

This matters more than it used to. Gartner's research found that 61% of B2B buyers prefer a rep-free buying experience. Whatever you make of that number, the direction is clear: more of the decision happens while you are absent. Your proposal is often the only salesperson in the room.

The numbers behind proposals that win

Good data on proposals is rare, because few companies can see large numbers of them. The largest public dataset we found comes from Proposify, which analysed 742,137 proposals across 30 industries. Proposify sells proposal software, so read these as useful signals from one vendor's customers, not as universal laws.

Two cautions before you act on this. First, correlation is not cause. A shorter proposal might win because the seller understood the client well enough to leave things out, not because of the page count itself. Second, these are the results of one company's customers, who already chose to use proposal software. Track your own close rate, then use these numbers to ask better questions about it.

Even so, the shape of the evidence supports a simple view. Clear beats long, and people care a great deal about what the price buys.

The structure that works

You do not need a clever template. You need sections in the order a reader's questions arrive. Here is the structure we recommend, and it maps closely to that average of seven sections.

1. What we understand about your business

Start with them. In two or three short paragraphs, describe their business, their goal and the problem as they would describe it. If you have spoken with them, use their words. A sentence like "You told us the phone goes quiet every January" does more work than any claim about your experience.

This section has one purpose: to make the reader think, "They were listening." Nothing else in your proposal will work as well once that thought is missing.

2. What we found

This is where most proposals are weakest and where you can pull furthest ahead. Do not say a business "could improve its online presence." Show something specific and checkable: they rank below three named competitors for the search that matters most, their mobile pages are slow, or their conversion tracking appears to be missing.

Specific findings do two things. They prove you did the work before asking for money, and they let the prospect verify your claim in a minute, which builds more trust than any testimonial you could write yourself.

3. What it is costing them

A problem without a cost is an observation. Connect each finding to something the owner cares about, such as calls, bookings or sales. Be careful here. If you are estimating, say that you are estimating and show how you got there. Inflated numbers are the fastest way to lose a careful reader, and a careful reader is exactly who signs.

Use language such as "Based on the search volume for these terms, we estimate" rather than "You are losing." If you cannot reasonably put a number on it, describe the consequence in plain words instead.

4. The plan

Now, and only now, talk about what you will do. Keep it to phases that a non-marketer can follow. Name what happens first, what the client needs to provide, and when they should expect to see progress. An owner who can picture the next ninety days is far more likely to commit than one facing a list of unfamiliar services.

Describe outcomes alongside tasks. "Fix the broken pages so that visitors from your Google listing reach a working booking form" lands better than "Technical remediation."

5. Proof that you have done this before

Pick two or three pieces of relevant work, not your whole portfolio. A roofing company wants to see other trades or local service businesses. It does not need your campaign for a software startup. Include a real result where you have permission to share one, and a real quote if you have one. Never invent or round up either.

6. Your investment

Price gets its own section, and it deserves some care. We cover it in detail below.

7. Next steps

End with exactly what happens next. Tell them how to accept, what happens after they do, and who will be in touch. A proposal that ends on a price leaves the reader alone with their doubts. One that ends on a clear next step gives them something to do.

How long should a marketing proposal be?

As short as it can be while still doing its four jobs. The Proposify data points to about 11 pages for winners against about 13 for losers. That gap is not large, and you should not trim a page just to hit a number. It does, however, push back against a common instinct: adding more to seem more thorough.

Here are three practical tests for length:

• The phone test. Could an owner get the point while standing in a car park, reading on a phone?
• The skim test. If someone reads only the headings and the price, do they still understand your offer?
• The removal test. If you delete a paragraph, would the client miss it, or only you?

Size also depends on the deal. A local business on a modest retainer usually wants something short and direct. A larger company with several decision makers may need more detail, particularly around scope and process. Match the document to the size of the decision.

How to present your pricing

Buyers spend real time on pricing, a reported 7.72 minutes on average in the Proposify data. That is the moment your proposal is most exposed, so think carefully about what they see there.

Offer options, but not too many

Offering one price turns the decision into yes or no. Offering a few turns it into "Which one?" For most agencies, two or three packages work well. Fewer feels like an ultimatum, and more makes the choice tiring.

Make each option clearly different, and describe who it suits. "Growth, for businesses ready to compete across their whole area" helps a reader place themselves. Then recommend one openly. You are the expert, and clients generally want your opinion.

Show what the price buys

Never leave a number standing alone. Put the deliverables next to it in plain language, so the reader can answer the question they will soon be asked: "What are we actually getting?"

The figures in that table are for illustration only. Your prices should reflect your costs and your market.

Be honest about what you cannot promise

Nobody controls Google. If you promise first place, a careful buyer will suspect you are either inexperienced or dishonest, and they may be right. Promise the work, the reporting and the standards you control. Share what similar work has achieved, with honest caveats. Paradoxically, this tends to increase trust.

Design: why looks matter more than you would like

Agencies sell judgment and taste, whether they like it or not. If your proposal looks careless, a prospect will reasonably assume your work will too. That does not mean you need an elaborate design. It means consistency: your own logo and colours, readable type, generous spacing and pages that work on a phone.

A few things matter more than people expect:

• Mobile first. Many owners open a proposal on their phone. A PDF built for A4 forces them to pinch and zoom, and some will not bother. A responsive live page is easier to read.
• Real pictures of real work. One honest screenshot of your results is worth more than ten stock photos of people pointing at laptops.
• A personal cover. Their name on the cover sends a small but clear signal that this was made for them.

Personalise the parts that matter

Personalisation has a bad reputation because too much of it is cosmetic. Dropping a company name into a template fools nobody. Real personalisation is in the findings, the plan and the choice of proof.

The tension, of course, is time. Researching a prospect properly used to take hours: running reports, checking Google Maps by hand, testing the site's speed, taking screenshots, moving it all into a document and pricing it. So agencies face a bad choice between pitching fewer prospects with care or more prospects with generic templates.

This is the exact problem Pitchlens was built to solve. You paste a prospect's website, it audits search, Google Maps, website health, speed, tracking and ads, then it builds a branded proposal from what it found, using your logo, your past work and your packages. You review and edit everything before it goes out. It does not replace your judgment, but it does remove the copy and paste, so your time goes on the parts that actually require you.

Using AI without losing trust

AI is now a normal part of writing proposals, and it can be excellent for a first draft. It is also capable of stating false things with total confidence. In a document with your name on it, that is a serious risk.

If you use AI, set some firm rules:

• Never let it invent numbers. Every statistic, result and price should come from your data.
• Never let it invent clients or results. A fabricated case study could end the relationship, and possibly your reputation, if discovered.
• Read every word. You are responsible for what goes out, not the model.
• Keep your voice. AI drafts tend to sound alike, and your prospect has probably read hundreds of them.

We take this seriously in our own product. In Pitchlens, AI can write words around your data, but any draft that introduces a number not found in the audit or your packages gets rejected automatically. That rule exists because one invented figure can undo the trust that the rest of the document has earned.

Sending it: link or PDF?

Send a live link as the main version, and attach or offer a PDF as a backup. Each has a job. The link works on any phone and can tell you when it has been read. The PDF suits anyone who needs to forward it, print it or file it.

Write a short covering message. Remind them of the one thing you most want them to remember, and be clear about the next step. "The main thing I'd look at is the Maps section on page three. Can we talk on Thursday?" is better than "Please find attached our proposal."

Better still, walk them through it live when you can. Presenting a proposal yourself lets you read reactions and handle questions as they come up.

Following up without being annoying

A surprising number of deals are lost simply because nobody followed up well. Waiting and hoping is not a strategy. Neither is sending "just checking in" every three days.

Timing matters. In a 2011 Harvard Business Review study, researchers concluded that "most companies are not responding nearly fast enough" to online leads. That research looked at inbound leads rather than proposals, and it is now old, but the underlying idea has lasted: interest fades, so reach people while your work is on their mind.

This is where knowing when your proposal is read changes everything. If you can see that someone opened it this morning and spent a while on pricing, your call is no longer an interruption. It is well timed.

Here is a simple follow-up rhythm that you can adapt:

1. Same day. Send a short note confirming it has been sent and restating the next step.
2. Soon after it is first opened. Wait a little, then offer to answer questions. Do not mention that you saw them open it, as it can feel intrusive.
3. After a few days with no reply. Share one useful thing, such as a competitor's move or a quick win they could make themselves.
4. After a week or so. Ask a direct question: "Is this still a priority, or should I check back next quarter?"
5. Close it politely. If you hear nothing, say you will assume the timing is wrong and leave the door open. This often prompts a reply.

Each message should give them something rather than ask for something. That is the difference between following up and chasing.

Common mistakes that lose deals

Most lost proposals share a handful of avoidable problems.

A checklist before you press send

Before any proposal goes out, read through this list:

• Does the first page talk about them rather than you?
• Is there at least one finding they could check for themselves?
• Are all numbers real, sourced and labelled when they are estimates?
• Could a non-marketer follow the plan?
• Is the proof relevant to their kind of business?
• Is it clear what each price includes?
• Have you recommended an option?
• Does it read well on a phone?
• Is the next step obvious?
• Do you know when you will follow up?

If you can answer yes to all ten, your proposal is already better than most of what your prospect will receive.

Turning this into a repeatable system

One excellent proposal is good. A process that produces excellent proposals every week is what actually grows an agency. Here is how to get there:

1. Save your building blocks. Keep your packages, past work and standard process in one place so that you are not rewriting them.
2. Audit before you write. Always start with evidence about that specific business.
3. Define your standard. Decide your structure once, so that every proposal your team sends meets the same bar, whoever writes it.
4. Track your close rate. You cannot improve what you do not measure. Note which sections people read and where deals stall.
5. Review your losses. When you lose, ask why. Most prospects will tell you if you ask politely, and those answers are worth more than any benchmark.

That system is what we built Pitchlens around: audit, proposal and tracking in one flow, under your own brand. You can see every feature here, compare plans, or read our guide on using an audit to win new clients.

Frequently asked questions

What should be included in a marketing proposal?

A strong marketing proposal covers what you understand about the client, what you found, what the problem is costing them, your plan, relevant proof, your pricing and clear next steps. Those seven sections answer the questions a buyer asks, in the order they ask them.

How long should a marketing proposal be?

As short as it can be while still making your case. In Proposify's analysis of over 700,000 proposals, winning proposals averaged about 11 pages and losing ones about 13. For a small local business, shorter is often better. Larger deals with several decision makers can justify more detail.

How many pricing options should a proposal include?

Two or three usually works best. A single price turns the decision into a yes or no, while too many options make people put off choosing. Describe who each option suits and recommend one.

Should I send a marketing proposal as a PDF or a link?

Use both if you can. A live link is easier to read on a phone and can show you when it has been opened. A PDF is useful for anyone who needs to forward, print or file it.

When should I follow up on a proposal?

Send a short note on the day you send it, then follow up soon after it has been read, if you can see that, or within a few days if you cannot. Each follow-up should offer something useful rather than just asking for an update.

Can I use AI to write a marketing proposal?

Yes, for drafting, as long as you check every word and never let it invent numbers, clients or results. Keep your statistics, prices and case studies tied to real data, because you are responsible for what goes out under your name.

Sources

• Proposify: sales closing statistics from 742,137 proposals. Figures are from the vendor's own customer data.
• Gartner: 61% of B2B buyers prefer a rep-free buying experience
• Harvard Business Review: The Short Life of Online Sales Leads