AI Wrote My Proposals
I spent 4 hours writing every proposal and sent them 5 days after the discovery call. My close rate was 24%. Now AI drafts every proposal in 45 minutes and it is sent the same day. My close rate is 41%. The proposals did not get worse — they got better, faster, and more consistent than I could ever manage manually.
The Counter-Intuitive Reality
The assumption most people make when they hear AI writes the proposals is that AI-generated proposals must be generic — the kind of copy-paste template that clients can tell was not written specifically for them. This assumption is wrong, for a specific reason: the personalisation in a good proposal comes from the discovery call insights, not from the writing. The discovery call reveals the client’s specific situation, their specific goals, and their specific concerns. If those insights are captured accurately in the debrief, the AI proposal is more specific to the client’s situation than the average manually-written proposal — because the AI systematically incorporates every detail from the debrief rather than the ones the writer happened to remember.
The second reason AI proposals work: they arrive the same day. A proposal received while the client is still processing the discovery call conversation lands in completely different psychological territory from one received 5 days later. At the 5-day mark, the urgency has subsided, 2 or 3 competitor calls have happened, and the prospect is in comparison mode rather than the motivated mode they were in immediately after the call. Same-day proposals close at 2 to 3 times the rate of delayed proposals — a fact so consistent that it has become the single most impactful change we recommend to any service business.
What Happens Between Call and Send
Immediately after the call: write the debrief
The moment the discovery call ends — literally within 5 minutes — open a notes document and write for 10 minutes without editing. Cover: the client’s situation in their own words (what specific phrases did they use to describe their problem?), their stated goal and the specific outcome they want, the timeline they mentioned, any budget signals (even indirect ones — their reaction to your pricing range), the concerns or objections they raised, what makes this project unique or more complex than a standard engagement, and the key decision criteria they mentioned. This 10-minute debrief is the most important input — everything else is template and AI.
Generate the proposal with the AI prompt
Pass the debrief and the proposal template prompt from Post 214 to Claude. The prompt instructs Claude to: write the executive summary in the client’s language (mirroring their specific phrasing from the debrief), frame the situation section around their stated problem and goal, describe the proposed approach in terms of outcomes rather than activities, build the deliverables list from the scope discussed in the call, and structure the investment section with clear justification rather than just a number. The first draft arrives in under 3 minutes.
Review and personalise in 20 minutes
Read the draft from the client’s perspective: does every paragraph reference something specific about their situation, or are there sections that could apply to any client? The generic sections — usually the company overview and the standard methodology description — get upgraded with specific client references. Add the one personal touch the AI cannot know: a specific example from your experience that directly addresses their situation, or a reference to a specific moment from the call that showed you were genuinely listening.
Format and send within the same working day
Paste the reviewed draft into the proposal template (PandaDoc, DocuSign, or a formatted PDF). Add your logo, the client’s logo if you have it, and any relevant case study pages. Send with a 3-sentence covering email: what you discussed on the call (one sentence to confirm you listened), what the proposal contains (one sentence to set expectations), and the single clearest next step (book a call to discuss or sign and let’s get started). The proposal arrives while the call is still fresh. The client feels seen. The close rate improves.
📌 The most common proposal mistake that AI does not fix: a proposal structure that is about you rather than about the client. Most proposals follow the structure: company overview, our services, our team, our process, investment. The client-centred structure is: your situation, your goal, our proposed approach for your specific situation, what you will receive, the investment, and why us. AI follows whichever structure you give it — make sure your template is client-centred before the AI drafts from it.
What if the client asks for a proposal immediately at the end of the call?
This is the ideal scenario — the motivation is highest at this moment. Say: absolutely, I will have it to you by end of day today. Then write the debrief the moment the call ends and generate the proposal while the details are fresh. A proposal sent 3 hours after the call closes at an even higher rate than one sent the same day but hours later. The client who asked for it immediately is signalling strong intent — reward that intent with the fastest possible response.
Do clients know the proposal was AI-assisted?
There is no obligation to disclose AI assistance in proposal writing — just as there is no obligation to disclose that you used a spell checker, a proposal template, or a copywriter. What matters is that the proposal accurately reflects the scope discussed, the pricing agreed, and your genuine commitment to deliver. If those are true, the proposal is yours — AI helped you write it more clearly and more quickly than you would have done manually. The expertise, the relationship, and the accountability are irreplaceably yours.
Want AI Proposal Generation Built for Your Business?
SA Solutions builds proposal generation systems — discovery call templates, AI drafting workflows, branded proposal formats, and e-signature integration.
