How to Write an Agency Proposal That Closes (2 Pages, 3 Options)

Quick answer: A proposal is a confirmation, not a pitch. If the client is reading it to decide whether they want you, you sent it too early. Keep it to two pages: their problem in their own words, what you will do, what it costs, and how to say yes. Send it within 24 hours of the call, give three options rather than one, and follow up if you have heard nothing in three days.

Most agency proposals are long because the writer is nervous. Twelve pages of methodology, team bios and process diagrams are a way of justifying a price to someone who has not yet agreed there is a problem worth solving.

The proposals that get signed are short, because the selling already happened. Here is how to write one, and what has to be true before you do.

The rule that decides everything

Never send a proposal to someone who has not verbally agreed there is a problem worth paying to fix. A proposal cannot create belief. It can only write down a belief that already exists.

Before you write a word, you should be able to answer three things from the call:

• What is the problem, in their words? Not "poor conversion rate" but "people ring up and say they could not book online."
• What is it costing them? Ideally a number they gave you, not one you calculated at them.
• Who signs, and when? If you do not know this, you are writing to a spectator.

If any of those is missing, you do not need a proposal. You need another conversation.

The structure, in order

• 1. The problem, quoted back. Open with their situation in their own language, ideally with the exact phrase they used. This is the section that makes them feel understood, and it is the only one most clients read carefully.
• 2. What it is costing. One short paragraph tying the problem to money. "Roughly 60% of your visitors arrive on a phone, and the booking form does not work on one."
• 3. What you will do. Plain deliverables, not methodology. "Rebuild the booking flow. Fix mobile speed. Set up call and form tracking so we can see what changes."
• 4. What it costs. Three options, priced. See below.
• 5. Timeline. Start date, milestones, finish. Realistic, with room in it.
• 6. What you need from them. Logins, content, sign-off points. This one quietly prevents most project delays, because it makes their side of the work visible before it becomes an excuse.
• 7. How to say yes. One clear instruction. "Reply with the option you want and I will send the agreement and invoice today."

Two pages. If it needs more, the scope is not clear enough yet.

Three options, not one

A single price is a yes or no question, and no is always the easier answer. Three options change the question to "which one", which is a far better question to be arguing about.

• Essential. Solves the one problem they actually raised. Cheapest, and genuinely useful on its own. Nothing crippled.
• Recommended. The one you would choose, priced roughly double, solving the problem plus the obvious next thing. Say plainly that this is your recommendation and why.
• Complete. Everything, including ongoing work. Some clients take it, and its real job is to make the middle option look measured.

Most people take the middle one, which is why the middle one should be what you actually want to sell. For how to arrive at the numbers themselves, see our guide to pricing agency work.

Language that costs you deals

• "I think we could probably help with..." Hedging reads as doubt. Say what you will do.
• "Leverage our proprietary methodology." Nobody buys a methodology. They buy more bookings.
• Deliverables with no outcome attached. "20 hours of SEO" is a cost. "Service pages for your three highest-value treatments, so you appear when someone searches for them" is a result.
• An open-ended scope. "Ongoing optimisation as needed" is how projects eat their own margin. Say what is included and what is not.

Send it fast, and follow it up

Within 24 hours of the call, while the conversation is still real to them. A proposal that arrives a week later is competing with a version of the discussion they have already half forgotten.

Then chase it, because silence on a proposal is almost never a decision. Three working days later: "Any questions on the proposal? Happy to walk through it, or adjust the scope if the middle option is more than you need right now." That last clause matters. It invites a smaller yes instead of forcing a no.

If two follow-ups go unanswered, close it out the same way you would end a cold sequence: one short, pressure-free message that makes it easy to say "not now". Our follow-up guide covers why that message is so often the one that gets answered.

Attach the evidence

The strongest proposals come with the audit stapled to them. It shifts the document from your opinion to a record of their business: measured mobile speed, the booking paths that do or do not exist, their rating against the practices around them.

It also does something subtle. It proves the problem existed before you turned up, which is the difference between a proposal and a sales pitch.

Tellsign generates that report in your own branding, so you can send it alongside the proposal with no mention of us. For the audit itself, see our 10-point checklist, and if you are still filling the pipeline that leads to these calls, start with finding clients on Google Maps.