BW-201a · Module 3

AI-Assisted Proposal Drafts — Speed for Structure, Judgment for Persuasion

4 min read

AI is genuinely useful for proposal writing, in specific and bounded ways. Used correctly, it accelerates the structural scaffolding — the outline, the section headers, the first-draft prose that establishes the flow — and leaves the human writer free to focus on the elements that determine whether the proposal wins: the specific problem description, the evidence selection, the pricing frame, and the final prose pass that gives the document its voice.

Used incorrectly, AI produces proposals that are structurally sound and persuasively hollow. They read as if they were written for every possible client, because they were. The AI has no knowledge of the client's actual situation, their specific pain, their decision-making context, or the conversation you had in discovery. It has the template. You have the intelligence. The winning proposal is what happens when the template meets the intelligence.

Do This

  • Use AI to generate the structural outline: section sequence, headers, approximate length per section
  • Use AI to draft the solution and methodology sections where specificity is lower
  • Use AI to check prose clarity, reduce passive voice, and identify redundant sections
  • Apply your discovery intelligence to write the problem statement and evidence sections — these require human knowledge

Avoid This

  • Ask AI to write the problem statement — it will write a generic version that could apply to any client
  • Use AI-generated proposals without a full 3-pass revision
  • Let AI select the proof elements — it will choose what sounds good, not what is most persuasive for this reader
  • Trust AI to set the pricing frame — that requires knowledge of the client's decision context that only you have

A practical AI workflow for proposals: Start with your discovery notes and a prompt that specifies the client's situation, the problem you are solving, the solution approach, and the key proof points. Ask the AI for a structural outline and a first-draft executive summary. Review the executive summary for specificity — it will almost certainly need to be made more specific and more pointed. Then use the outline to write (or prompt AI to draft) the body sections, applying your specific intelligence at each step. Run the 3-pass revision on the result.

The AI-assisted proposal takes roughly the same amount of total revision time as a human-drafted proposal — the time savings come in the drafting phase, where AI eliminates the blank-page problem. What AI cannot replace is the editorial judgment that determines whether the document is actually persuasive, which is the only judgment that matters.

One specific caution: AI proposal prose has a recognizable texture — smooth, authoritative-sounding, and slightly generic. Readers who evaluate many proposals are increasingly sensitive to it. The tell is usually in the problem statement (too abstract) and the solution description (too complete — AI does not know what to leave out). A strong prose pass in the voice of the writer you actually are will address this. If the document does not sound like you at all, it will not sound credible to the reader who met you.