Which tool to choose
The tool is the conversational assistant, used to build the argument: you explain who you're proposing to, what you're offering and what objections you expect, and it gives you a structure that leads the reader toward yes. For the numbers — costs, expected revenue, return on investment — the spreadsheet remains the tool for the math; AI writes the proposal around figures you've verified. For recurring proposals, keep a template. The rule: AI to persuade and structure, the spreadsheet for the numbers, your judgment for how much to promise.
How to do it
- Explain to the AI the recipient, what you're proposing, the problem it solves and what doubts the decision-maker will have.
- Have it give you the structure: summary, problem, solution, benefits, costs, why choose you.
- Enter the real numbers and verify them in the spreadsheet; don't let the AI estimate them at random.
- Anticipate the objections in the text and reread with a critical eye: does the proposal promise things you can deliver?
A concrete example
Antonio had to propose a ten-thousand-euro service to a client and didn't know how to frame it so as not to seem merely expensive. He explained to the AI the client, the problem and the value, and asked for a convincing proposal structure. The assistant built it for him: the cost of the current problem, the solution, the expected return, why him. Antonio put in his real numbers, verified in the spreadsheet, and refined the tone. The proposal no longer talked about price but about value, and the client understood why it was worth it. AI had given the persuasive form; he gave the solid numbers.
When it does NOT work (and how to fix it)
If the numbers don't add up or are too optimistic
AI can get the math wrong or inflate the benefits to make the proposal more attractive. Do the math yourself in the spreadsheet and keep the estimates realistic: a proposal that promises too much backfires when the results don't come. The numbers must hold up to the decision-maker's scrutiny.
If it sounds like a generic proposal
A proposal that could apply to anyone convinces no one. Give the AI the specific details of the client and their problem, and ask it to talk about that. The more the proposal is tailored to the recipient, the better it works.
If it ignores the decision-maker's objections
The reader will have doubts — it's too expensive, what if it doesn't work, why you. A good proposal anticipates and dissolves them. Tell the AI which objections you expect and have it address them in the text, instead of leaving them as unspoken doubts.
A tip from someone who really uses it
Shift the conversation from price to value, and that's where AI helps most. A weak proposal lists what you do and how much it costs; a strong one first shows how much the problem the client already has is costing, and then presents the expense as an investment that pays off. The assistant is good at building this persuasive structure, as long as you give it the real elements of the case. But never delegate the numbers to it: a proposal lives or dies on the credibility of the figures, and those must be yours, verified, sustainable. Persuasion from AI, math from the spreadsheet, promises you can keep from your judgment: that's the combination that gets a yes.
Frequently asked questions
Can AI calculate the quote for me?
It can help you frame it, but do the exact math in the spreadsheet and verify it: AI can get the sums or the estimates wrong. The figures of a proposal must hold up to the decision-maker's examination, so responsibility for the numbers is yours.
How do I make the proposal convincing and not generic?
By giving the AI the specific details of the client and the problem, and having it talk about those. A proposal tailored to the recipient, one that knows their context, convinces; an abstract one doesn't.
Should I include the objections?
Yes: the best proposals anticipate the decision-maker's doubts and dissolve them. Tell the AI which objections you expect and have it address them in the text.
If AI writes it well, is the proposal ready to send?
No, and it's the haste that loses clients or, worse, gets them accepted on terms you can't sustain. A well-written proposal with wrong numbers or inflated promises is a trap: it convinces today and collapses tomorrow, when the results don't come or the math doesn't add up. Before sending it, verify the figures in the spreadsheet and ask yourself whether you can really deliver what you're promising. AI gives you the persuasive form; the sustainability of numbers and commitments is the check that falls to you, and that separates a winning proposal from a postponed problem.