Yes, if the hour is spent assembling rather than writing. Build a proposal kit once: past proposals, approved case paragraphs, a pricing menu and standard terms. Then spend each hour in five blocks: turn call notes into a brief, draft from the kit, set scope and price yourself, have AI critique the draft as the buyer would, then run a final check.
Two caveats keep this honest. The first proposal done this way takes three or four hours, because you're building the kit as you go; the under-an-hour result usually arrives by the third or fourth. And AI speeds up the parts of a proposal that are writing, not the parts that are judgement. It can't tell you whether this client is a good fit, what the engagement is worth to them, or what you can deliver in the days you're pricing. Those take the 15 minutes in the middle of the hour, and they should.
The proposal kit you build once
Put these in a shared Project (ChatGPT and Claude both offer Projects on business plans) or a folder you can reference from Copilot in Word. Expect about half a day to assemble a first version.
- Three to five past proposals you won, anonymised: client names, figures and identifying details removed. Reusing past client work without leaking client data covers how.
- A service menu: each engagement type you sell, with typical scope, deliverables and duration.
- Case paragraphs: short, approved accounts of past work, each with only the results you can evidence.
- A pricing menu: day rates, fixed-fee ranges for standard packages, and your rules for discounts.
- Standard sections: your approach, team biographies, assumptions and exclusions, terms.
- A style note: length, tone, words you never use, how you structure options.
Anonymising a won proposal takes more than deleting the client's name, because small details identify a client to anyone in the same sector. One paragraph from the operations consultancy's brokerage proposal, before it went into the kit and after:
| As sent to the client | As stored in the kit |
|---|---|
| Your five-person renewals team, led by your operations director since the merger last spring, handles around 1,400 renewals a year through your broking system. | The client's renewals team of four to six people handles a four-figure number of renewals a year through their broking system. |
The merger, the leader's role and the exact renewal count would each point a competitor or a former employee to one firm. Size bands and sector are enough for the model to reuse the paragraph's shape, and they're all a new prospect needs.
The style note is the kit document people write most vaguely. A filled-in one for the same consultancy runs to six lines: proposals of 900 to 1,200 words; second person, addressed to the decision-maker; the client's problem in their own words in the first paragraph; three options with the middle one recommended; days shown next to every fee; never "bespoke", "end-to-end", "best-in-class" or "journey". Each line is something the model can follow or break, which makes breaches easy to spot in the check.
Here's a filled-in case paragraph for an illustrative two-person operations consultancy. Note what it leaves out:
CASE: Workflow review, 12-person insurance brokerage
Problem: renewals handled in three different ways across the
team; frequent late re-quotes.
What we did: 6-day review; mapped the renewal process with the
team; agreed one standard process and checklist; set up renewal
reminders in their existing broking system.
Result (evidenced): all renewals on the single process within
two months; client's own count of late re-quotes fell over the
following quarter.
Do NOT claim: a percentage improvement (client didn't measure
one), cost savings.The "Do NOT claim" line is there because models embellish case studies. Asked to write about this project, a model will happily say "reduced late re-quotes by 40%", a figure that exists nowhere. That's the fastest way to lose a buyer who checks references. Catching made-up figures in AI-drafted proposals covers the checks.
The hour, block by block
| Minutes | Block | Who does it | Output |
|---|---|---|---|
| 0-10 | Call notes into a one-paragraph brief | AI drafts, you correct | A brief you'd be happy to read back to the client |
| 10-25 | First draft from the kit | AI | A full draft with placeholders for scope and price |
| 25-40 | Scope, price, assumptions | You | Two or three options with fees and exclusions |
| 40-50 | Critique as the buyer | AI objects, you fix | A list of weak points, then a revised draft |
| 50-60 | Final check and send | You | A proposal you'd sign |
Minutes 0 to 10: from discovery-call notes to a one-paragraph brief
A good brief is the difference between a proposal that sounds like it was written for this buyer and one that could have gone to anyone. Paste your call notes, however rough:
From these discovery-call notes, write a one-paragraph brief:
the client's situation, the problem in their words, what they
want to be different in six months, constraints (budget, timing,
people), and who decides. Use only what's in the notes; mark
anything missing as [ASK]. Then list three questions I should
confirm before sending a proposal.
Notes: {paste}An illustrative result for a recruitment agency prospect:
"A 14-person recruitment agency placing temporary warehouse and driving staff. Consultants spend much of Monday rebuilding timesheets from texts and emails, and invoice errors are causing disputes with two large clients. The managing director wants timesheets 'done by Monday lunchtime without anyone chasing' within six months. Budget: [ASK]. Timing: wants to start before the peak season in November. Decision-maker: the managing director, with the operations lead's input. Questions to confirm: 1) What timesheet system, if any, is in use now? 2) Is there a budget range? 3) Who owns the client relationships affected by invoice errors?"
One correction: the notes said peak season starts "in the autumn", not November; the model sharpened a vague phrase into a month. Correct it, or confirm it with the client, because it drives your timeline. The [ASK] on budget is right, and the questions are useful; send them in a short email while you draft.
That email works best short and specific, so the client can answer from their phone. For the agency:
"Thanks for the time today, [first name]. I'm putting the proposal together now and want to get three things right. Which timesheet system, if any, are your drivers using at the moment? Is there a budget range you'd like me to work within, even a rough one? And who looks after the two clients where invoices have been disputed? I'll have options with you by [day] either way."
Sending it during the hour rather than after means answers often arrive before the proposal goes, and a budget range in particular changes which options you offer.
Minutes 10 to 25: a first draft from the kit, not from the internet
The instruction that matters is where the content comes from:
Using the brief above and only the documents in this Project
(service menu, case paragraphs, standard sections, style note),
draft a proposal with: summary of their situation; what we
propose; approach and timeline; relevant experience (use case
paragraphs verbatim, don't add results); team; next steps.
Leave [SCOPE] and [PRICE] placeholders for me.
Under 1,200 words. Follow the style note.In Copilot in Word, you can do the same by typing "/" in the Copilot prompt box to reference specific files, such as your service menu and a past proposal; drafting proposals from your own files in Copilot shows it step by step.
The biggest difference is in the opening. Before, a draft without the brief and kit tends to start: "In today's competitive market, efficient operations are essential for recruitment agencies seeking to grow." After, with them: "Your consultants spend most of Monday rebuilding timesheets from texts and emails, and two of your largest clients are disputing invoices as a result. You want timesheets finished by Monday lunchtime without anyone chasing." The second version tells the managing director you listened. It came straight from the brief.
Watch the experience section even with "verbatim" in the prompt. In a realistic slip, the draft rewrote the brokerage case paragraph as "a workflow review for a similar-sized staffing business", tailoring the case to the prospect. It reads better and it's untrue: the client was an insurance brokerage. A buyer who asks for a reference in staffing would find out quickly. Compare the experience section against the kit, word for word, in the final check; if the model has "improved" a case paragraph, paste the original back in.
Minutes 25 to 40: scope, price and assumptions stay with you
This block is yours. Decide what you'll do, in how many days, for what fee, and what's excluded. Offering two or three options helps buyers say yes to something. A filled-in options table for the agency:
| Option A: Diagnose | Option B: Diagnose and fix | Option C: Fix and support | |
|---|---|---|---|
| What's included | Process review, findings report, recommended timesheet approach | A, plus configuring the chosen timesheet tool and training the team | B, plus 8 weeks of fortnightly check-ins through peak season |
| Days | 4 | 9 | 12 |
| Fee | [your price] | [your price] | [your price] |
| Excludes | Software licences; changes to client contracts | Software licences; payroll changes | Software licences; payroll changes |
The fees come from your pricing menu, not the model. Illustratively, if the consultancy's menu says a day rate of $1,100, fixed fees rounded up to the next $250, and a 10% contingency on any option that includes configuring software, the options price themselves: A is 4 days, $4,400; B is 9 days plus 10%, $10,890, rounded to $11,000; C is 12 days plus 10%, $14,520, rounded to $14,750. Writing the rule down once means every proposal prices the same way, and the consistency check below can test the arithmetic as well as the words.
Once you've filled it in, AI is useful as a consistency checker rather than a pricer: "Check that the days, deliverables and exclusions are consistent across these options and with the draft; list any mismatch." It will catch the draft promising "training for all consultants" when Option A doesn't include training, which is the kind of slip that becomes a dispute.
An illustrative response on the agency draft listed three items. First, the approach section promises training for all consultants, but Option A excludes training. Second, the timeline section describes six weeks of check-ins for Option C, while the table says eight. Third, Option A excludes changes to client contracts but B and C don't mention them. The first two are real mismatches and get fixed. The third is deliberate, since B and C include work that may touch client contracts, but it's a fair sign the exclusions need a sentence of explanation rather than a bare list.
Minutes 40 to 50: let the AI argue as the buyer
You are the managing director of the agency in the brief. You've
received three proposals this week. Read this one and list the
five things most likely to make you hesitate or choose someone
else, most serious first. Be specific and blunt. Don't rewrite it.An illustrative response, shortened:
"1. You haven't said what happens if the timesheet tool you recommend doesn't suit my drivers, most of whom only use their phones. 2. Option B's nine days don't show how much of my team's time you need; I can't spare consultants in October. 3. Your experience section is about an insurance brokerage, not recruitment. Why should I trust you with temporary staffing? 4. No mention of the two clients disputing invoices, which is my most urgent problem. 5. The next steps ask me to sign before we've agreed a start date."
Points 1, 2 and 4 are worth fixing in ten minutes: one line on mobile-first tools, a line on the team's time commitment, and a sentence addressing the invoice disputes directly. Point 3 is fair but you can only answer it honestly: explain the parallel (a process run three ways by a busy team) rather than inventing recruitment experience. Point 5 is a matter of preference; many consultants keep a signature request in the proposal. Treat the critique as a list of candidates, not instructions.
Minutes 50 to 60: the send check
- Every name, figure and date matches the brief and the call notes.
- No leftover detail from a past proposal: search for previous clients' names and sectors.
- Case paragraphs say exactly what the kit says, with no added results.
- Scope, days, deliverables and exclusions match across the draft and the options table.
- No placeholder left: search for "[", "ASK", "SCOPE", "PRICE".
- The first paragraph is about them, not you.
If you send through proposal software such as PandaDoc, Proposify or Qwilr, you also get e-signature and view tracking. Whether that's worth a per-seat subscription over a PDF and a general assistant depends on volume; AI proposal software versus a general assistant works through it.
One proposal, timed end to end
Here's how the hour went, in illustrative figures, for the two-person operations consultancy on the recruitment agency proposal above, its fourth using the kit.
- Brief: 9 minutes. One correction (the November date) and three questions emailed to the client.
- Draft: 12 minutes. Generated in under two minutes; the rest spent reading it through and swapping one case paragraph for a better match.
- Scope and price: 21 minutes, six over plan. Deciding whether Option C was worth offering took the time, and it was the right place to spend it.
- Buyer critique: 11 minutes. Three fixes made, as above.
- Check and send: 8 minutes. Found a sentence referring to "your renewals team", left over from the insurance brokerage proposal in the kit.
Total: 61 minutes. The consultancy's previous proposals for similar work took about three hours, most of it writing sections that are now in the kit. Those numbers are illustrative, but the pattern is typical: the writing blocks shrink dramatically, the thinking block doesn't, and the check block catches something every time. Leftover text from an old proposal is the most common catch, which is why "search for previous clients' names" is on the list.
A compliance consultant who works with financial planners ends up with a heavier kit and a lighter hour. Their engagements are more standard (annual file reviews, policy updates), so the service menu does most of the drafting, and the hour often comes in closer to 40 minutes. The buyer critique still earns its place, usually by pointing out regulatory deliverables the draft describes vaguely.
When a proposal shouldn't be squeezed into an hour
- Formal tenders with scored questions, mandatory formats and compliance statements. Use AI for drafting answers from your kit, but plan the time the tender needs.
- Your first proposal for a new service. There's no kit to draw on yet; build one as you write it.
- High-value or strategic bids, where a day spent on the proposal is small against the fee.
- When the brief is still unclear. If the [ASK] list is long, a second call beats a fast proposal. The questions from minute 10 are often the best thing to send first.
For the preparation that makes the brief good in the first place, see how consultants use AI for client research before discovery calls, and for what most often sinks AI-written proposals, the AI proposal mistakes that cost you the job.
Faster proposals with AI: follow-up questions
Should I tell the prospect that AI helped write the proposal?
There's no general duty to, but be ready to say how you work if asked, and make sure nothing in the proposal claims a personal effort it didn't get. What buyers care about is that the diagnosis, scope and price are yours and correct. If your future contract will involve AI tools on their data, say so in the proposal's approach or terms.
Can AI tell me what to charge?
Not reliably. It has no view of your costs, your pipeline or what this buyer will pay, and its price suggestions are guesses dressed as benchmarks. Use it to check your pricing logic instead: ask it to find inconsistencies between the scope, the days estimated and the fee, or to list what a buyer might think is missing at each price option.
Does this work for a new consultancy with no past proposals?
Yes, with a slower first month. Write the kit from what you have: your service descriptions, two or three anonymised examples of past work from employment or projects you can describe, a pricing menu and your terms. Your first three proposals build the rest. Keep the best sections of each as you go, and the kit fills up quickly.
Further reads
- The AI Tool Stack a One-Person Consultancy Actually Needs — The tools a one-person consultancy actually needs around this.
- How Much Time Can AI Save a Solo Consultant Each Week? — Realistic weekly time savings for a solo consultant.
- Claude vs ChatGPT for Business Writing, Proposals and Reports — Which assistant writes proposals and reports better.
- How to Write Business Proposals Faster With AI — The general proposal method outside consulting.
- How to Mention AI Use in Client Contracts and Proposals — How to mention AI use in contracts and proposals.
- How to Set Up Claude Projects as a Shared Team Assistant — Set up a shared Project to hold your proposal kit.
- How to Write Event Proposals With AI in Half the Time — Where the hours in an event proposal really go, the kit to build once, four prompts to run in order, and a timed worked example.
- How Small Agencies Use AI to Write New-Business Pitches — Where AI shortens a small agency's pitch: go/no-go scoring, prospect research, idea stress-tests, case-study matching and a sceptical-buyer rehearsal.
- How Freelancers Use AI for Invoices, Proposals and Chasers — One thread from proposal to payment: AI drafts the scope, checks the invoice and writes the reminders, and your invoicing software sends them on time.
- Automate Client Onboarding With AI: Forms, Contracts, and Emails — One intake form, a template contract filled from it, and emails triggered by the signature, with AI writing the kick-off brief and chasing what's missing.
- AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.
Sources: Microsoft Support, Draft and add content with Copilot in Word; OpenAI and Anthropic pages on Projects for business plans; PandaDoc, Proposify and Qwilr product pages.