Quote generation: from enquiry to a priced quote your team approves.
Quote generation reads an enquiry, asks for any missing details, applies your price rules and fills your quote template. A person checks and approves each quote before it goes out. The AI handles reading and writing; the prices come from rules you control.
How does an enquiry become a quote?
The slow part of quoting is rarely the arithmetic. It is reading a vague enquiry, working out what the customer actually wants, asking the questions that were not answered and then typing it all into a template. A quote generator does those steps in order:
- Reads the enquiry from email, a web form or WhatsApp, and extracts the job: what, where, how much, by when.
- Asks for what is missing, such as measurements, quantities or photos, before pricing anything.
- Applies the price rules to produce line items, using a price list and rules you maintain.
- Fills your template with the line items, terms, validity period and a short cover note.
- Sends it for approval to the person responsible, who edits if needed and releases it.
- Records it in your CRM or accounting system so it can be followed up and converted to an invoice.
Is quote generation a job for an agent?
Tick what is true of your quote generation work.
Tick the statements that apply and the suggestion appears here: an AI agent, ordinary automation, or a person with software support.
What price rules need to be written down?
If your prices live in one person’s head, the first job is getting them out. That exercise alone often shows inconsistent pricing between staff.
| Input | Example |
|---|---|
| Price list | Unit rates for materials, products or services, kept in one table. |
| Labour or time rules | Hours per task, minimum call-out, rates for evenings or weekends. |
| Modifiers | Distance or zone, access difficulty, rush jobs, volume discounts. |
| Tax | Which VAT rate applies to which line. Irish VAT rates differ by type of supply, so confirm the rules with your accountant. |
| Terms | Validity period, deposit, payment terms, exclusions. |
| Limits | Jobs too large, unusual or risky to quote without a site visit. |
Why prices come from rules, not the AI
A language model is good at understanding “we need the back bedroom and landing done, ceilings too, it’s a 1970s semi”. It should not decide what that costs. The model turns the enquiry into structured quantities; ordinary code multiplies them by your rates and adds tax. That way the same job always gets the same price, and you can see exactly how a total was reached.
Where the model does add value is in everything around the price: spotting that the customer mentioned a deadline, noticing that a requested service usually needs a second one, writing a cover note that refers to what they actually asked, and listing assumptions in plain words. Those are the parts that make a quote read as if someone paid attention.
How much time could quote generation give back?
Put your own numbers in.
The starting values are an illustrative example, not a benchmark or a neu figure. Use your own volume, the time each one takes today and the share you expect the system to handle without a person.
Time saved is only part of the case: response times, missed work and errors matter too. The ROI guide covers how to set a baseline, and the cost guide what the build and running costs include.
Where the quote lives
Accounting tools such as Xero and QuickBooks can store quotes or estimates and convert them to invoices, which keeps numbering and tax consistent. A CRM such as HubSpot tracks quotes as part of a deal. Check which object your plan supports through the API before deciding where quotes are created.
For trades and other site-based work, photos and measurements matter. The generator can ask for them by WhatsApp and attach them to the job, so the person approving the quote sees what the customer sent.
What can go wrong
- Quoting from too little information. The generator should prefer a question to an assumption, and say in the quote what it assumed.
- Stale prices. Supplier costs change; someone needs to own the price list and update it on a schedule.
- Jobs that need a visit. Some work cannot be priced remotely. The rule should be to book the visit, not guess.
- Quotes that are never chased. Pair it with follow-up automation so sent quotes get a timely nudge.
- Approval bottlenecks. If quotes wait days for the one person who can approve them, speed gained elsewhere is lost. Name a backup approver and set limits each person can sign off.
How to measure it
Track time from enquiry to quote sent, how often the approver changes the price, and the share of quotes that convert. If approvers keep adjusting the same kind of line, the rule is wrong and should be fixed at source. Many businesses start by combining this with lead qualification, so only serious enquiries get a full quote. For cost, see the AI agent cost guide.
Keep every quote and every approver change. Over a few months, that history shows which job types you consistently underprice, which customers accept quickly and where the price list needs revising. It is also a good dataset for testing any change to the rules before it goes live.
Common questions
Can it quote for custom or one-off jobs?
Only as far as your rules cover them. For genuinely bespoke work, the generator can still save time by collecting the details, drafting the scope and pulling similar past quotes for reference, while a person sets the price. Mark those quotes clearly so they are not mistaken for rule-based ones.
Will customers see the quote is automated?
They see a quote on your template, approved by a member of your team, so in practice it is your quote. If the conversation leading up to it happened with an automated assistant, tell customers that at the start, along with how to reach a person.
Can it offer discounts?
Only discounts you define, such as volume or repeat-customer rates, and preferably with a ceiling. Discretionary discounts should stay with a person. Any discount applied should appear as its own line, so the approver and the customer can both see it.
What if our prices are in a spreadsheet?
That is a common and workable starting point, as long as the sheet is laid out as a proper table. The generator reads it on each quote, so updating the sheet updates the quotes. Lock editing to the people responsible for pricing.
Bring us the task.
Tell us what happens today, which tools are involved and where it goes wrong. On a free 30-minute review, we’ll work out whether an agent, an automation or neither is the right fix, and what a sensible first release would be.
Start a project.

Book a call or send us a message. We read everything.
tash@neu.ie
Working with teams in Ireland and worldwide.
Free 30-minute AI review.