To write a proposal with AI, start from the request (an RFP, a brief, or notes from the client call), let an AI proposal writer draft against it, then check the draft against every requirement and review it with your team. AI saves the most time on the first draft and on rewording. The part that wins proposals, understanding exactly what the reader asked for, is the part to keep a close eye on.
What a Proposal Needs
- Summary: the problem, your approach and the outcome, in a paragraph
- Understanding of the need: the request in your own words, so the reader knows you heard it
- Approach and scope: what you will do, and what is out of scope
- Timeline and milestones
- Team: who does the work and why they are suited to it
- Budget or pricing
- Next steps: what the reader should do to go ahead
Writing a Proposal with AI in Fable
1. Bring in the request
Create a project and import the RFP or brief (PDF and Word files both work, with headings and tables kept), along with a past proposal you are proud of. See import and export.
2. List what was asked
Ask the project assistant: "List every requirement and question in the RFP, with its section number." Keep the answer; it becomes your checklist.
3. Draft it
Start a new document from the Project Proposal template, or ask the assistant to draft the proposal as a new document based on the RFP and your notes, and approve it. You can also open a blank document and type or speak the instruction: "Write a proposal to redesign the client's onboarding flow. Three phases over ten weeks, a team of four, fixed price. Match the structure of the RFP."
4. Check it against the request
Ask: "Which requirements from the RFP does the proposal not address yet?" Then fill the gaps with targeted instructions: "Add a section on accessibility testing in response to requirement 3.4." Each edit shows what changed. See editing with AI.
5. Review with the team and send
Invite colleagues as Editors for their sections, or as Viewers who leave typed or spoken suggestions. The version history records every change and who made it, so you can restore any earlier version. Export to PDF to submit, or DOCX if the client wants an editable copy.
Tips for Better Proposals, With or Without AI
- Use the reader's words. If the RFP says "member portal," do not call it "customer dashboard."
- Answer in their order. Following the request's structure makes your proposal easy to score.
- Be specific about outcomes. "Reduce onboarding time" is weaker than "cut onboarding from three weeks to one."
- Say what is out of scope. It prevents disputes later.
- Check names, numbers and dates last. They are the errors readers notice first.
When to Use Which Tool
If your team sends many similar sales proposals and needs a content library, pricing tables, CRM integration or e-signatures, dedicated proposal software is built for that. Fable is for writing the proposal itself, especially longer, custom ones that several people contribute to and that go through rounds of revision. For more on collaborative writing, see the best collaboration tools for writers and editors, and if the RFP arrives as a PDF, how to edit a PDF with AI.
Frequently Asked Questions
Can AI write a proposal?
AI can draft a strong first version of a proposal from the request, your notes and past proposals, and it is good at structure and clear language. Your team still decides the approach, the price and the commitments, and checks that every requirement is answered.
How do I use an AI proposal writer with an RFP?
Import the RFP, ask the AI to list every requirement and question in it, draft the proposal against that list, then check the draft against the list before you submit. In Fable, the RFP and your draft can sit in the same project so the assistant can compare them.
Does Fable have a proposal template?
Yes. New documents in Fable offer a Project Proposal template you can start from, and you can then fill and revise it by typing or speaking instructions.
Can several people work on the same proposal?
Yes. Invite colleagues to the project as Editors to write sections, or as Viewers who leave suggestions for the owner to accept or reject. Every change is saved to the version history with its author.