1. Read the request
AI extracts the need, deadline, context, attachments and unclear points from the email or form.
Blog · Small business guide · AI quotes · Claude + n8n
Automatic AI quotes: the practical workflow for small businesses
An automatic quote system should not sell in your place. It should read the request, prepare a clean draft from your price list, ask for approval and follow up without forgetting.
Short answer
What AI should automate, and what it should not decide
The right setup prepares most of the work. It does not change your prices, make promises for you or send sensitive quotes without approval.
2. Draft the quote
It applies your price list, uses your template and writes the lines in plain business language.
3. Follow up
If the prospect does not answer, it prepares a short follow-up with the right context.
Workflow
The concrete workflow
Incoming request
Client email
A prospect describes the job, often with missing details.
Claude summarizes the request and lists what is missing.
You get a clean brief before opening the quote.
Draft
Document ready
n8n pulls the model, price list and client data.
AI drafts the quote lines and the email.
You review instead of starting from a blank page.
Follow-up
Tracked answer
The system records the send date and reply status.
After 5 days, it prepares a personal follow-up.
You approve, then automate simple cases later.
Guardrails
Limits to set from day one
Pricing
AI does not decide your margin
It applies approved rules. If a discount is outside the frame, it stops and asks you.
Data
No good quote without a clean base
Your offers, options, deadlines and conditions need to be written somewhere first.
Control
Validate everything at first
A sales automation should start as an assistant, then become a co-pilot, then handle simple cases.
Payback
The realistic gain for a small business
If you handle 10 requests a week and each quote takes you 30 minutes, that is 5 hours of repetitive work. A good circuit brings preparation down to 5 or 10 minutes per quote. The real gain is not only the time: your answers go out faster and follow-ups no longer depend on your memory.
It is not worth it if you put out two bespoke quotes a month, highly complex and heavily negotiated. There, a library of templates and a drafting assistant are enough.
FAQ
Frequent questions
Can an automated quote still be personal?
Yes, if the system starts from your real templates, your tone and your commercial rules. The point is not to send a generic quote faster.
Will it work with the software I already use?
Often yes, if your tool accepts exports, emails, APIs or automations through Make or n8n. Otherwise we start with a semi-automatic version.
Do I need to code?
No, not on your side. The technical part is installed with you, on your own accounts, so you can use it without becoming a developer.
Next
Where to go next
If your quotes still go out by hand, the circuit described above is the one I build in custom AI agents.
The AI audit
Ninety minutes on your tasks, and you leave with a plan ranked by priority. The step to take when everything looks urgent at once.
AI training for your team
Four private sessions so the skill stays in-house. Your first systems get built during the training, not after it.
Custom AI agents
I install the system on your own accounts, with your rules. You review, you approve, and the task runs without you.