GEO Proposal Template for Freelancers and Agencies

#01What makes a GEO proposal win?

Clients rarely choose the longest or cleverest proposal. They choose the one that shows the provider understood their problem, offers a clear plan and feels low-risk. In GEO, where many buyers are unsure what they need, clarity matters even more.

Most losing proposals share the same weaknesses: generic problem statements, vague deliverables, long explanations of how AI works, and promises nobody can keep. Our view: if a client can't tell from your summary page exactly what they'll get, when and for how much, the rest of the proposal won't save it.

#02What sections should the proposal include?

Copy the structure below. Replace the italic guidance with your content for each client.

Section 1: Summary

One page. Restate the problem in one or two sentences, using the client's language from your discovery call. Describe the proposed work in three or four bullet points. Give the total price, timeline and start date. End with the decision you need, such as "Approve by Friday to start on the 1st."

Section 2: What we heard

Show you listened. Summarise the client's situation: why they're looking at AI answers, which buyer questions matter, which assistants they think their buyers use and what a good outcome looks like. Include one or two quick observations you made before the call, such as a screenshot of an inaccurate answer. Keep it factual.

Section 3: Scope and deliverables

Be specific. List every deliverable with quantities. For an audit: prompt set size, engines, runs per prompt, technical checks, source review, recommendations report and presentation. For a sprint: number of page rewrites, new pages, listing corrections and the re-test. For a retainer: monthly tracking, report format and capacity.

Add a short "Out of scope" list for common requests that aren't included, and a "Client responsibilities" list: access, approvals, subject-matter experts and review turnaround.

Section 4: Method

Keep it brief. Explain how you'll build the prompt set, why you run each prompt several times, how you record results and how you prioritise recommendations. One or two paragraphs is enough. Clients want confidence you have a method, not a lecture.

Section 5: Timeline

Show the milestones. Kick-off, data collection, analysis, draft findings, final presentation. For sprints, include review rounds and the re-test date. Note dependencies, such as "Page rewrites start once the client provides CMS access."

Section 6: Measures

Say how results will be judged. Name the measures, such as mention rate by question type, accuracy of descriptions, citations of the client's pages and AI referral visits. Explain that you'll use the same prompts and method at the start and end, so results are comparable. Be clear that you can't control how assistants respond and won't guarantee specific outcomes.

Section 7: Price and terms

Make it simple. State the fixed price or monthly fee, what it covers, payment schedule, how extra work is quoted and how either side can end a retainer. If you offer options, limit them to two or three, described by deliverables.

Section 8: About you

Short and relevant. Two or three sentences on your experience in the client's category, one relevant case study and, if you have them, a testimonial or reference. Skip long company histories.

Section 9: Next steps

Make it easy to say yes. Explain how to approve, such as signing the attached agreement or replying by email, and what happens in the first week after approval.

#03How do you write the problem statement?

Use the client's own words. If they said "the AI keeps recommending Competitor A," write that, not "suboptimal generative visibility." Connect it to something they care about: pipeline, brand accuracy, sales conversations or competitive position. Keep it short; one strong sentence beats a paragraph.

#04How do you describe deliverables clearly?

Use nouns and numbers. "A 30-prompt baseline across three assistants, each prompt run three times" is clear. "Comprehensive AI visibility analysis" isn't. For each deliverable, state the format, such as a spreadsheet, a report or a presentation, and when the client receives it.

#05How do you present options?

If you offer more than one option, make them build on each other: an audit alone, an audit plus sprint, or an audit plus sprint plus three months of retainer. Describe the differences in deliverables. Recommend one and say why, briefly. Too many options slow decisions; three is the usual maximum.

#06How do you handle guarantees?

Don't offer guarantees about mentions, citations or rankings. Instead, guarantee what you control: deliverables, deadlines, method and honest reporting. If a client asks for a guarantee, explain that nobody controls assistants' answers, and offer a review point where they can stop if results don't justify continuing. That offer usually satisfies the concern behind the request.

#07How long should the proposal be?

Three or four pages for most projects, with the summary on page one. Put detailed prompt lists or example reports in an appendix if needed. Send it as a PDF or shared document the client can open easily, with your name, the client's name and the date in the file name.

#08From MentionX data: why proposals should name enginesOriginal data

Method. Read-only aggregate queries over MentionX scans from 22 August to 29 October 2026 (UTC): 2,217 engine runs (2,154 completed) across 18 anonymised projects and six engines.

MentionX's anonymised aggregates show how much engines differ, which is why a proposal should state which ones you'll test:

Our view: naming the engines and run counts in the scope section prevents later disagreement about what "AI visibility" meant.

Limits. A convenience sample of mostly B2B software projects, run through engine APIs, which can behave differently from the consumer apps. Perplexity, Copilot and DeepSeek figures come from a single day (26 October 2026), so treat them as a snapshot, not a trend.

#09Worked example: A proposal for Ledgerline

Ledgerline is a fictional accounts-payable tool. Figures are illustrative.

A freelancer sends Ledgerline, a fictional accounts-payable tool, a four-page proposal the morning after a discovery call. Page one says: "Sales reps keep hearing that AI assistants recommend Competitor A for invoice-approval questions." It proposes a three-week audit: 30 prompts, three assistants, three runs each, a technical check, a source review and a presentation, at a fixed price with half paid upfront.

The scope lists what's out: content writing and developer work. Client responsibilities include a product expert for one hour a week. The measures section explains that results will be compared using the same prompts. The marketing director forwards page one to her CFO, who approves it the same day.

#10How do you follow up after sending?

Send the proposal with a short email: what's attached, the one-line summary and when you'll follow up. If you haven't heard back by that date, send a brief, friendly note asking whether they have questions. Offer a short call to walk through it. Don't chase repeatedly; two follow-ups are usually enough.

If they ask for changes, respond quickly with a revised version, and change the scope rather than just the price. Keep a dated version history so everyone knows which proposal they're approving.

#11How do you reuse the template without sounding generic?

Keep the structure fixed and rewrite sections two and three every time. The client's situation and the specific deliverables should never be copied from another proposal. Save a library of method paragraphs, timelines and terms you can reuse, but read each one in context before sending. Clients notice another company's name or a mismatched detail immediately, and it costs trust you can't easily win back.

#12What should you check before sending?

Read the proposal as the client's boss would. Can you tell what's being bought, for how much and why, in under two minutes? Check that every number is right, every deliverable has a quantity and every promise is one you can keep. Confirm the client's name, dates and contact details. Then send it as a single file with a clear name.

Finally, keep a copy of every proposal with its outcome. After a dozen, you'll see which sections clients ask about, which options they choose and where your scope tends to need tightening. That record is the best way to improve your proposals over time.

#13How do you adapt the template for agencies?

Agencies usually need a few additions. Name the team members who'll work on the account and their roles, so the client knows who they'll deal with day to day. Explain how the account will be managed: a main contact, a regular check-in and an escalation route. If you subcontract any work, such as technical fixes, say so.

Larger clients may also ask for security, data protection or procurement information. Keep a standard pack ready with your policies, insurance details and any certifications, so you can respond quickly without rewriting the proposal. Put it in an appendix rather than the main document.

#14How do you adapt it for a retainer renewal?

A renewal proposal is shorter. Summarise what you delivered in the last period, show results against the agreed measures using the same method, and propose the scope for the next period. Highlight one or two lessons and how they shape the next phase. If you're changing the price or scope, explain why in a sentence. Clients renew more readily when the proposal reads like a continuation of a story they already know.

Send renewals at least a month before the current period ends. That gives the client time to approve without a gap in tracking, which would break the trend you've built together over the previous months. A gap of even one month makes it much harder to show progress.

#15Common questions

How long should a GEO proposal be? Three or four pages, with a one-page summary first.

Should I include a guarantee? Guarantee deliverables, deadlines and reporting, not mentions or citations.

How many options should I offer? Two or three at most, building on each other.

What should the summary include? The problem, the work, price, timeline and the decision you need.