AI · September 12, 2026 · 4 minute read

They hung up. The proposal was already written.

Basim Mousilli · September 12, 2026

AI

They hung up. The proposal was already written.

# They hung up. The proposal was already written. You had a good call. They talked for thirty minutes about what is broken, you asked the right questions, and at the end they said "send me something." Then Thursday happened. Then Friday. By the time you sat down to write it, you were reconstructing the call from three lines of notes and your memory of what they seemed to care about. The proposal you sent on Monday was not the proposal you could have written on Tuesday afternoon, and they had four days to cool off in the middle. That gap is where deals go. Not the pitch, not the price. The gap. ## What the delay actually costs you The proposal you write three days later is worse in a specific way: it is generic. You remember the shape of the conversation but not the sentences. The client said "I am the bottleneck on every quote" and you write "streamline your sales process." They said it better than you did, and you lost the line. So the first thing to fix is not speed. It is capture. ## Step 1: Stop taking notes Record the call and let something else transcribe it. There are several tools that join a Google Meet or Zoom call, record it, and produce a summary; we use Fathom's free tier. The tool matters less than the rule: you do not take notes during a discovery call. You listen. What you get afterwards is not a summary you skim and delete. It is the client's own words about their own problem, which is the single most valuable input you have for everything you will write to them from here. ## Step 2: Send the summary back before you send anything else Same day, while the call is still live in their head. Email or WhatsApp, no attachment, no pitch: > Here is what I heard. Tell me what I got wrong. This does two things. It proves you listened, which is most of what they were checking during the call. And it gives them a low-stakes way to correct you before the correction costs you a rewrite. Nobody does this. It takes four minutes. ## Step 3: Draft from the transcript, not from memory Paste the transcript into whatever you use to draft, with a prompt that fixes the structure so you are not re-deciding it every time. The one we use: > Create a detailed proposal based on the following client meeting summary. Structure it with sections for Background, Summary, Scope, Deliverables, Timeline, Payment Terms, and Get Started. Speak in plain English for a business owner with limited technical knowledge. Use the client's own phrasing where they described their problem. > > [transcript] The last line is the one that matters. Without it you get agency language back. With it, the Background section reads like the client wrote it, because they did. **Here is the part most posts about this leave out.** The draft is a draft. It gets the structure, the recall and the scope right, and it gets the judgement wrong. Pricing, what you are willing to promise on a timeline, and whether you want this client at all are yours. If you send the first output without reading it, you will eventually promise something you cannot deliver, and you will have done it in writing. Budget ten minutes to read it as if you were the person receiving it. ## Step 4: Make it easy to say yes on the same page The proposal should be a link, not a PDF, and the approval and the payment should live on that link. Every extra step between "yes" and money is another day for the yes to decay. A countersigned PDF that needs printing, an invoice that arrives separately tomorrow, a "let me know and I'll send payment details." Then the follow-up runs without you: the kickoff call gets booked, the reminders go out, the notes from the call sit somewhere you can find them in six months. Whatever you use for this, the test is whether it happens when you are not thinking about it. ## Where this does not apply If your proposals are competitive bids scored against a spec, or your scope genuinely takes engineering input to price, this is not your process. Speed is not the constraint there; being right is. This is for the service business where the discovery call is the sales call, the scope is a conversation, and the only thing standing between a good call and a signed deal is that you have twelve of them and no time to write. --- *Basim Mousilli is the Creative Director at Insert Fuel.* If you want help setting this up in your own business, [book a call](/book-a-call).

Put AI to work properly

AI is only worth it when it does a real job.

You have just finished “They hung up. The proposal was already written.”. There is a lot of noise about AI right now and very little of it survives contact with a real business. We use it where it genuinely earns its place: answering leads at midnight, drafting the follow-up you never get round to, and building pages and small apps in an afternoon. Let us show you the parts that would actually help you.

  • Real jobs it can do in your business this month
  • The AI Vibe Builder comes with every plan
  • Straight answers about what it cannot do yet
See the fuel plans

Thirty minutes, free, and you will speak to Basim rather than a salesperson. If we are not the right fit we will say so and point you somewhere better. Not ready for a calendar? Leave us a sentence instead and we will ring you.

Previous ArticleYou said the smartest thing all month on a client call, and nobody else heard itMarketingNext ArticleYour consulting site says less than your pitch deck does