Draft proposals and quotes in ChatGPT without inventing price

What you will be able to do

You will draft customer proposals and quotes in ChatGPT from prices and scope you already approved, using a two-pass draft/tighten method so ChatGPT does not invent line items.

Ground price and scope first

Before you open ChatGPT:

• Pull the rate sheet or ranges you stand behind
• List inclusions and exclusions you already use
• Note deposit, cancel, and validity rules that already exist
• Write the customer’s real scope in bullets (what was asked on the call)

If a number is not approved, it does not enter the prompt as a “guess.” It enters as [NEED] for you to fill.

Two-pass method

Pass 1 — Draft from grounded facts
“Using only the facts below, draft a clear proposal: summary, scope, exclusions, price table placeholders, next step. Do not invent prices or discounts. Mark gaps [NEED].”

Pass 2 — Tighten
“Cut filler. Keep every fact. One idea per short paragraph. Put the ask in one sentence.”

Same sandwich as email drafts and the Claude proposals twin: model in the middle, you on both ends.

Use a Quotes Project

Keep approved rate rules and a winning anonymized sample in a Quotes Project. Today’s customer scope goes in the chat. Standing files stay stable.

Optional custom GPT or Canvas worksheet

For staff who quote often, build a quote helper with blank rate cells. Still copy finals into your real proposal path.

Desk check before send

• Every price matches the approved sheet or a manager override you recorded
• Exclusions match policy
• No surprise discount or “standard package” ChatGPT invented
• Validity date and deposit language match your real terms
• One clear next ask (approve, schedule site visit, reply yes/no)

What ChatGPT should not own alone

Final contract language, regulated clinical claims, and anything “sign this” without your review process. ChatGPT drafts the narrative; you own the commitment.

Try this job: rewrite one real quote narrative

1. Pick a quote you already sent (strip identifiers if training staff).
2. Paste only approved rates + that job’s scope bullets.
3. Run Pass 1 and Pass 2.
4. Diff against the sent version: did ChatGPT add soft extras?
5. Save the prompt + checklist for the next estimator.

Short close

Approved price and scope in, two passes, desk check—ChatGPT drafts; you commit.