You finished the site visit hours ago. You know the job, you know roughly what to charge, and the customer is waiting. But the quote is still not written, because writing it means sitting down, wording every line, remembering your terms, and making it look professional instead of like a text message. So it slides to tonight, then to tomorrow, and by then a competitor who quoted the same day is already booked.

Learning how to write a quote for a job is not really about math. You already know your prices. The slow part is turning rough job notes into a clean, priced document that reads like a business the customer wants to hire. That is the part AI is genuinely good at. Not setting your prices, and not deciding your margins, but taking the numbers and details you give it and shaping them into a tidy quote in minutes.

Below is a simple system: a reusable ChatGPT or Claude workspace, a briefing that teaches it your business once, and copy-paste prompts that turn field notes into a finished quote. If you have never built one of these workspaces, start with our walkthrough on ChatGPT Projects for small business, then come back here for the quote-specific prompts. And once the quote is out the door, our guide on how to follow up on a quote with AI covers the polite nudges that turn a sent quote into a booked job.

What a Good Quote Really Contains

Before you hand anything to AI, it helps to know the four parts of a quote that keeps you out of trouble. A quote that is just a name and a big number invites questions, haggling, and “but I thought that was included.” A clear quote answers those before they are asked.

Scope of work. Plainly what you will do, in the order you will do it. “Remove old fence, dispose of debris, install 80 feet of six-foot cedar privacy fence with two gates.” This is what protects you later.
Line-item pricing. Break the job into parts the customer can understand: materials, labor, disposal, permits. You do not have to itemize every screw, but a few clear lines beat one mystery total.
Assumptions. The quiet source of most disputes. Note what you are assuming: access to water and power, no hidden rot, work done in a single visit, customer clears the area first.
Terms. Deposit, payment schedule, how long the price holds, and what happens if the scope changes. A simple “50% deposit to schedule, balance on completion, price valid 30 days” removes a lot of awkward conversations.

You supply the substance. AI handles the wording, the structure, and the polish, so every quote you send hits all four without you writing them from scratch each time.

A Quick, Honest Note on Pricing

AI can make a quote look professional. It cannot tell you what to charge, and you should never let it try. If you ask a chatbot to “price a fence job,” it will invent numbers based on averages from who-knows-where, and those numbers will not know your costs, your area, your crew, or your margins.

So the rule for this whole workflow is simple: you bring the prices, AI arranges them. Feed it your real rates and quantities. Tell it to leave a clearly marked blank for any number you have not given it, instead of guessing. That one instruction is the difference between a tool that saves you time and a tool that quietly undercharges you. It is the same discipline we use when we automate customer emails with AI: the AI drafts, but it flags anything it does not know rather than making it up.

Step 1. Build One “Sales and Quotes” Project

A Project is a permanent workspace inside ChatGPT or Claude. You brief it on your business once, and every chat you start inside it remembers that briefing. Instead of re-explaining your trade, your pricing style, and your terms every single time, you open the Project, paste today’s job notes, and get a quote that already sounds like you.

You can build this on a free account. As of July 2026, the ChatGPT Projects feature is available on both the free and paid plans, and free users can upload up to five reference files per Project, while Plus and Go users can upload up to 25, per Coursera’s guide to ChatGPT Projects. Five files is plenty for a quoting workflow.

Prefer Claude? It works the same way for this job. As of July 2026, both tools have a free plan and their main paid plans run about $20 a month, per ChatGPT’s pricing page and Claude’s pricing page. Paying mostly buys higher usage limits, not a different feature here. If you are still deciding which tool to use, our guide on how to choose the right LLM for your business walks through that first.

If you already built a “Sales and Quotes” Project for following up on quotes, use that same one. Writing the quote and chasing the quote belong together.

Step 2. Brief It in Your Own Voice (and Your Own Numbers)

This is the part that makes your quotes sound like your business instead of a generic template. In the Project’s custom instructions box, brief the AI the way you would train a new estimator. Here is a block you can copy and swap in your own details:

I own a residential fencing company in Boise, ID.
Our customers are homeowners who want tidy, durable work and a
clear, no-surprises quote. Typical jobs run $2,000 to $12,000.

Brand voice: friendly, plain-spoken, professional. No jargon,
no hard sell. Read like a trustworthy pro who is easy to deal with.

You turn my rough job notes into a clean, professional quote.
Follow these rules every time:
- Structure every quote with: Scope of Work, Line-Item Pricing,
  Assumptions, and Terms.
- Use ONLY the prices, quantities, and rates I give you.
  Never invent, estimate, or "look up" a price. If a number is
  missing, insert a [BRACKETED BLANK] for me to fill in.
- Keep the scope specific and in plain English.
- My standard terms: 50% deposit to schedule, balance on
  completion, price valid 30 days. Include these unless I say otherwise.
- Keep it easy to skim. Short lines, clear headings, no fluff.
- Sign off as "Sam, Boise Line Fencing."

That bracketed-blank rule matters more than it looks. It stops the AI from filling a hole in your notes with a made-up figure, which is the fastest way a quote goes out too low. Then, if you have them, upload two or three reference files: a past quote you were proud of and your standard terms. Keep it lean. Three good examples beat thirty.

Step 3. Turn Rough Job Notes Into a Draft

Here is where the time saving shows up. Instead of writing a quote from a blank page, you paste in the messy notes you took at the job and let the Project shape them. Save this prompt inside your Project so you never rewrite it:

Turn my job notes below into a clean, professional quote using my
standard structure (Scope of Work, Line-Item Pricing, Assumptions,
Terms). Use only the numbers I provide. For anything priced that I
did not give you a number for, insert a [BRACKETED BLANK]. Do not
guess any price. Keep the scope specific and easy to read.

JOB NOTES:
[customer name, address, what they want, measurements, materials,
your labor and material prices, timeline, anything unusual]

Your notes can be as rough as you like. Something like: “Dupree, 1420 Oak. Wants privacy fence back and one side, about 130 ft, 6 ft cedar, two gates. Materials ~$3,100. Labor 3 days, $2,400. Haul old chain-link, $250. Start in 2 weeks.” The AI will turn that into a structured quote with clean headings, a readable scope, and your terms attached, and it will flag anything you left out.

Read what comes back like a customer would. Is the scope specific enough that there is no room for “I thought that meant…”? Are the numbers yours, with no surprises the AI slipped in? If a bracket is still sitting there, that is the AI doing its job: telling you it needs a number instead of inventing one.

Step 4. Tighten the Scope, Assumptions, and Terms

The first draft gets you most of the way. A few quick follow-up prompts inside the same chat make the quote airtight. You do not need to retype anything; the Project remembers the quote it just wrote.

Sharpen the scope so it protects you:

Rewrite the Scope of Work so it is specific enough to prevent
"I thought that was included" disputes. Spell out what IS and is
NOT part of this job based on my notes. Do not add work I did
not mention.

Surface the assumptions you forgot:

Based on this job, list the assumptions I should state so I am
covered if conditions are different than expected (access, site
prep, hidden damage, single visit, weather). Only list ones that
fit this job. Keep them short and plain.

Offer a good-better option, only if you want one:

Draft an optional "good and better" version: the quote as-is, plus
one upgrade the customer might want, priced separately using a
[BRACKETED BLANK] for the upgrade cost. Keep it low-pressure.
Do not invent the upgrade price.

That last one is optional. Some trades win more work by showing a simple upgrade; others just want one clean number. Use it when it fits, skip it when it does not.

Step 5. The 60-Second Check Before You Send

AI gets you most of the way. The last stretch is a quick human pass, and on a quote it is the most important 60 seconds of the whole job, because this document becomes what you agreed to. Before any quote leaves your hands, run this check:

Fill every bracket. If the draft still says [PRICE] or [DEPOSIT], you are not done.
Check the math and the numbers. Add the line items yourself. Confirm the total is right and the prices are yours.
Confirm the scope matches the job. Right address, right work, right materials.
Cut anything you cannot stand behind. AI sometimes adds a warranty line or a discount to sound helpful. Delete anything you did not authorize.
Read it out loud once. If it does not sound like you, change one line.

Sixty seconds, then send. This is the same habit we recommend across every AI workflow: move fast, but never let the machine send an unread document a customer will hold you to.

Common Mistakes to Avoid

Letting the AI set the price. The single biggest risk. If your instructions do not forbid guessing, the AI will happily invent a number that undercharges you. Keep the bracketed-blank rule in every prompt.

A vague scope. “Fix the fence” is where disputes are born. The more specific the scope, the fewer arguments later. This is exactly what the Step 4 scope prompt is for.

No assumptions or terms. A quote without a deposit, a validity window, or a note on hidden conditions leaves you exposed. Build them into your Project once and they show up every time.

Sending it hours late. The quickest quote often wins, even over a slightly cheaper one. The whole point of this system is to send the same day, while you are still fresh in the customer’s mind.

Forgetting to follow up. A great quote still needs a nudge. Pair this with a follow-up rhythm so quotes do not die in silence.

Frequently Asked Questions

How do I write a quote for a job step by step?

Start with the four parts every quote needs: scope of work, line-item pricing, assumptions, and terms. Gather your real numbers first, your material costs, your labor, and any extras like disposal or permits. Then, instead of writing from a blank page, paste those notes into a ChatGPT or Claude Project briefed on your business and let it structure them into a clean quote. Finish with a 60-second human check: fill any blanks, add the math, confirm the scope, and cut anything you did not authorize. The AI handles wording and layout; you own the numbers and the final read.

Can AI figure out how much to charge for a job?

No, and you should not let it try. AI does not know your costs, your local market, your crew’s pace, or the margin you need to stay in business, so any price it invents is a guess dressed up as a number. Use AI for the wording and structure of the quote, and bring your own prices. The workflow here tells the AI to leave a clearly marked blank for any figure you have not provided, so it never quietly undercharges you. Pricing is a business decision that stays with you.

Is it free to write quotes with AI?

Yes, for most owners. As of July 2026, ChatGPT Projects works on the free plan, with up to five reference files per Project, per Coursera’s guide, and Claude also has a free plan. Paid plans on both tools run about $20 a month as of July 2026, per ChatGPT’s pricing page and Claude’s pricing page, and mostly raise your usage limits. Most owners start free and only upgrade if they hit a wall.

What is the difference between a quote and an estimate?

A quote is a fixed price you commit to for a defined scope of work. An estimate is your best-guess figure that can change once the real work begins, common when the full scope is not yet known. The AI workflow here writes both; you just tell it which one you are sending. If it is an estimate, say so plainly in the document and note what could move the number, so the customer is not surprised by the final bill.

How do I keep the quote from sounding robotic?

Brief the Project in your own voice, as shown in Step 2, and give it a real past quote you were proud of as a reference file. The AI matches the tone and structure of what you feed it. Then read the draft out loud before sending. If a line does not sound like something you would say, change it. The goal is a quote that reads like your business wrote it, because in every way that matters, it did.

Try This Today

Open ChatGPT or Claude and create one Project called “Sales and Quotes.” Paste the custom-instructions block from Step 2 and swap in your trade, your typical jobs, your standard terms, and your sign-off. Then grab one job you visited recently, type your rough notes and real prices into the Step 3 prompt, and read the quote that comes back. Fill any bracket, run the 60-second check, and send it.

That is under 15 minutes, and you walk away with a reusable workspace plus one real quote already on its way to a customer, instead of sitting in your notes app.

If you would rather not build this alone, that is what an AI coaching session is for. We look at how you really run your business, set up your quoting and follow-up workflow with you, and hand you the prompts, so you leave with a system that is working instead of a to-do list.

How to Write a Quote for a Job With AI (Clear, Priced, Fast)