Guide · Price & Quote
How to move from manual quoting to automated proposals
Most quotes are retyped from call notes into a template, then chased by memory. This guide walks through replacing that with a draft generated from the sales call, an approval rule, and an automatic follow-up, using tools you probably already have.
The short version
- Standardise your pricing and quote template
- Capture the scope from the sales call
- Turn the transcript into line items with AI
- Generate the quote document
- Route it for approval with discount guardrails
- Send it and automate the follow-up
- Hand the accepted quote to invoicing
1. Standardise your pricing and quote template
Automation can only copy what is already consistent. Before touching AI, write down every service you sell with its unit (hour, day, project, month) and price, plus the terms that go on every quote: payment terms, validity period, what counts as out of scope.
A rate card in a spreadsheet is enough. If you price projects from day rates, the free Rate Card to Quote Calculator shows the margin before the quote goes out.
2. Capture the scope from the sales call
Most of a quote is already said out loud on the call. Record and transcribe it with whatever your team already uses: Zoom, Google Meet or Teams transcripts, or a note-taker like Fireflies or Otter. Most of them can send the transcript to a webhook or an automation tool when the call ends.
If you do not record calls, a short structured form the rep fills in straight after the call (client, services discussed, quantities, start date) works almost as well and is easier to get approved.
3. Turn the transcript into line items with AI
This is the step that removes the retyping. Give a model like Claude the transcript and your rate card, and ask for structured line items, not a finished quote. Keep the maths out of the model: it extracts what was agreed, your quoting tool calculates totals and tax.
The prompt below is the pattern that holds up in practice. The two lists at the end, assumptions and out-of-scope, are what make the draft safe to send: they tell the rep exactly what to check.
You turn a sales call transcript into a draft quote.
Use only these services and rates (do not invent others):
<rate_card>
{paste your rate card: service, unit, price}
</rate_card>
From the transcript below, return JSON with:
- client_name, contact_email (if said)
- line_items: [{ service, quantity, unit, unit_price, note }]
Only include services the client asked for or agreed to.
If a quantity was not stated, set it to null and add a note.
- assumptions: things you inferred that the rep should confirm
- out_of_scope: anything the client mentioned that is NOT on the rate card
- payment_terms and start_date, if mentioned
- discount_requested: true/false, with the exact words used
Do not round, total or discount anything. The quoting tool does the maths.
<transcript>
{transcript}
</transcript>4. Generate the quote document
Map the JSON into your template. Options, from simplest to most connected: paste it into the free Quote Generator or Proposal Generator; fill a Google Docs template from an n8n, Make or Zapier workflow; or create the quote directly in your CRM or accounting tool (HubSpot quotes, QuickBooks or Xero estimates) through their APIs.
Give every quote a versioned number (Q-1042-v2). When the client asks for changes, you revise the same quote instead of starting a new one, and the accepted version is unambiguous.
5. Route it for approval with discount guardrails
Decide what can go out without a second look. A common rule: quotes under a set value with no discount send straight away; anything with a discount over 10%, a non-standard payment term, or an out-of-scope item goes to a manager first.
The approval can be a Slack message with Approve and Edit buttons. Checking the request against your rules is a good job for a structured AI decision: it returns approve, reject or escalate with a reason, instead of free text.
6. Send it and automate the follow-up
Speed matters more than polish: a quote that arrives the same day beats a better one that arrives next week. Once approved, send it from the rep's own email so replies land with them.
Then schedule the follow-ups at send time, not from memory: a check-in after 2 business days, a second after 5 with one useful detail (a start date, a question about scope), and a final one when the quote is about to expire. Stop the sequence the moment the client replies or accepts. An automation tool watching the inbox or the CRM deal stage does this reliably.
7. Hand the accepted quote to invoicing
The quote is the first billing document. When it is accepted, the same line items, payment terms and milestones should flow into the invoice or the billing schedule without anyone retyping them. That handoff is where most quote-to-cash leakage starts.
Which approach fits
| Approach | Fits when | Trade-off |
|---|---|---|
| Free generators + a template | Under ~20 quotes a month, one person quoting | Still typed by hand, but consistent and fast |
| Proposal software (PandaDoc, Proposify) | Design-heavy proposals, e-signature, tracking opens | Content is still entered manually |
| CPQ (DealHub, Salesforce CPQ) | Large product catalogues, complex pricing rules, sales teams of 20+ | Heavy setup and cost for small service firms |
| DIY: transcript + Claude + n8n/Zapier | Service firms quoting from calls, 20-100+ quotes a month | You maintain the workflow; the rep still reviews every draft |
Where this comes from
I built a scope-to-quote product on this exact flow before folding it into the free Quote Generator. The write-up and code stats are on the projects page.
Questions
What is quote automation?
Quote automation means the quote is assembled from information you already have, such as the call transcript, the CRM deal and your rate card, instead of a rep typing line items, prices and terms into a document. A person still reviews and approves it before it is sent.
Can AI write the whole quote?
AI is good at extracting what was agreed on a call and drafting the scope wording. It should not set prices or do the arithmetic: restrict it to your rate card, let your quoting tool calculate totals, and have a rep check the assumptions and out-of-scope items it flags.
How do I automate quote follow-ups?
Schedule the follow-up sequence when the quote is sent: a check-in after about 2 business days, another after 5, and one before the quote expires, and stop it automatically when the client replies, accepts or the CRM deal stage changes.
Do I need CPQ software?
Usually not for a service business with a short rate card. CPQ pays off with large catalogues and complex pricing rules. Most agencies and consultancies get further with a clean rate card, a template, and an automation that drafts the quote from the call.