A marketing agency proposal has one job: to make saying yes easy for someone who may not have been on the call.
Most proposals fail at that. They describe the agency, bury the price, leave the scope vague enough to argue about in month three, and end without telling anyone what happens next. This covers what to include, in what order, what to cut, and the follow-up sequence that recovers proposals that go quiet.
Before you write anything
Two decisions determine whether the document works, and both happen before you open a template.
Has the selling already happened? A proposal is a confirmation, not a pitch. If you are trying to persuade someone in a PDF, you have skipped a conversation. Send proposals only after a call where you heard the problem in their words.
Who else will read it? In most businesses the person you spoke to is not the only person deciding. Your document will be forwarded to a finance lead, a co-founder or a marketing manager who has no context. Write so that person can understand the offer without you in the room.
That second point is the single most common reason a good proposal dies. The champion understood it. The person holding the budget did not.
The structure that works
1. Their situation, in their words
Open with the problem as they described it. Not your founding date, not your list of services.
If they said "we get plenty of enquiries but half go cold before anyone calls back," that sentence goes near the top, close to verbatim. It proves you listened and frames everything after it as a response rather than a catalogue.
Two to four sentences. Resist the urge to demonstrate expertise here; the demonstration is that you repeated their problem accurately.
2. What you propose to do
State the work as outcomes, not deliverable names.
Weak: "SEO optimisation, content strategy, technical audit, monthly reporting."
Strong: "Fix the twelve pages that currently rank on page two so they earn clicks, and rebuild the enquiry form so leads reach you in under a minute instead of overnight."
The second is harder to write because it requires you to have diagnosed something specific. That difficulty is the point, and it is what separates a proposal from a price list.
Three to five items. More than five reads as padding.
3. Scope, stated as a boundary
This section prevents the argument you will otherwise have in month three. Write two lists.
Included. Specific and countable. "Four articles a month, 1,200 to 1,500 words, two rounds of revisions each." Not "content creation."
Not included. The section almost everyone omits, and the one that protects the relationship. "Paid ad spend, photography, video production, and changes to pages outside the twelve listed above."
An explicit exclusion is not negative or defensive. It is the difference between a scope conversation now, while goodwill is high, and a resentment conversation later, when it is not.
4. Timeline, with their obligations in it
Put dates against phases. Then, in the same table, put their dependencies: brand assets by the 5th, feedback within three working days, ad account access before week two.
Most missed agency deadlines are caused by the client and never documented. Writing it down does not make you difficult, it makes the timeline honest. It also gives you something neutral to point at when week four slips because access never arrived.
5. Price, presented as a decision
Three rules.
One recommended option. Choice paralysis kills deals. If you present tiers, present three, mark one as recommended, and make the middle one the recommendation.
Show the monthly figure and the total commitment. "$2,400 per month, three month minimum, $7,200 total." Clients work the total out anyway; burying it reads as a trick when they do.
State it flatly. No "we know this is a significant investment." Price, then next section. Apologising for your price teaches the reader that you are not sure about it.
6. What happens next
One sentence, one action. "Reply to this email to confirm and we will send the contract and a kickoff invitation the same day."
Ambiguity here is where proposals die quietly. A document that ends with "let us know your thoughts" gets filed under thoughts.
What to leave out
Company history. Nobody has ever bought because an agency was founded in 2017.
The full team page. Name the people who will actually do the work. Nobody else is relevant.
Generic case studies. One relevant example beats six irrelevant ones. If you have nothing in their industry, use a comparable problem and say so plainly. Pretending a fitness client is evidence for a law firm damages trust when they notice.
Methodology diagrams. The five-phase circular process graphic persuades nobody and signals the document was assembled from a template.
Terms and conditions in the body. Attach them. A proposal opening with liability language reads as defensive.
Length, format and delivery
Two to four pages. Any agency claiming their proposals run to twenty pages is describing a document nobody finishes.
Send a link, not an attachment. You find out whether it was opened, and you can fix a typo without sending a v2 that makes you look disorganised.
Plain formatting. A heavily designed proposal invites the reader to evaluate the design rather than the offer. It also raises an awkward question about what you spend your time on.
Pricing mistakes that lose deals
Hourly rates. The moment you price in hours you have invited a debate about how many hours something should take. You will lose that debate, because the client is imagining their own speed at a task they do not understand.
Line-iteming everything. Breaking a retainer into fourteen components creates fourteen opportunities to cut one. Package it.
Unexplained discounts. An unexplained discount teaches the client that your price is negotiable, which makes every renewal a negotiation. If you discount, attach a condition: a longer commitment, payment upfront, or a case study.
For context on what agencies actually charge, our guide to Facebook ads agency cost covers real market ranges, and the cost questions clients ask is worth reading before you get asked them live.
The follow-up decides more than the document
Most proposals are not rejected. They go quiet. The client got busy, the champion went on holiday, someone else needed to approve it, or a bigger fire started.
So sending is not the end of the process.
Day 0. Send it and state when you will follow up. "I'll check in Thursday if I haven't heard back."
Day 3. Follow up exactly as promised. Short. No new information, no discount, no anxiety.
Day 10. Second follow-up with one genuinely useful thing attached: something you noticed on their site, a competitor move, a relevant change in their market.
Day 21. Close the loop. "Assuming the timing isn't right, I'll stop chasing. Happy to pick this up whenever it is."
That last message recovers more deals than any other, precisely because it is the only one that does not ask for anything.
The practical difficulty is that nobody reliably sends messages three and four for eleven proposals at different stages. That is a tracking problem, not a discipline problem, which is why proposals belong in a pipeline stage with a next action date rather than in a sent-items folder. Agencies running several at once tend to keep them in a CRM built for agencies for that reason alone.
A skeleton you can copy
[Client name] - [one line naming the outcome]
THE SITUATION
Two to four sentences, in their words, about the problem.
WHAT WE PROPOSE
Three to five bullets, each an outcome rather than a task.
WHAT IS INCLUDED
Specific and countable.
WHAT IS NOT INCLUDED
The boundary. Two or three lines.
TIMELINE
Phase | Date | What we need from you, by when
INVESTMENT
[Amount] per month. [Length] minimum. [Total].
NEXT STEP
One sentence, one action.
Fill that in and you have something stronger than most twenty-page decks, because every line is doing work.
Frequently asked questions
How do you write a marketing proposal?
Start with the client's problem in their own words, then state what you will do about it as outcomes rather than tasks. Define what is included and explicitly what is not, give a timeline that includes the client's own obligations, present one recommended price with the total commitment shown, and end with a single clear next step. Write it so that someone who was not on the call can understand it, because that person often controls the budget.
What are five things all proposals should include?
A restatement of the client's problem, the proposed work expressed as outcomes, a scope boundary covering both what is and is not included, the price with its total commitment, and an unambiguous next step. Everything else is optional. Agency history, team pages and methodology diagrams are the most common padding and the first things to cut.
How long should an agency proposal be?
Two to four pages for most retainer and project work. Length correlates with lost deals rather than won ones, because long proposals get skimmed and the price becomes the only section anyone reads properly. If a proposal genuinely needs more than four pages, the scope probably needs another conversation first.
Should you put pricing in the proposal or wait?
Put it in, and qualify budget before you write the document. Sending a proposal without a price creates an extra round trip and signals uncertainty. If you are unsure whether the number will land, that is a sign the pricing conversation has not happened yet, and no document will rescue it.
What do you do when a proposal goes quiet?
Follow up on a schedule you stated when you sent it: roughly day 3, day 10, and day 21. Keep the first two short and add one genuinely useful piece of information to the second. The day 21 message should release them politely rather than ask again, which is what makes it the one that most often gets a reply.
Should agencies price hourly or by project?
By project or retainer, with the scope bounded. Hourly pricing invites a negotiation about how long the work should take, penalises you for getting faster, and caps your income at your available hours. The exception is genuinely open-ended consulting where nobody can define scope in advance.
