How to automate price quotes: a quote in minutes, not days

An enquiry arrives on Monday morning. The salesperson gets to it on Wednesday, sends the quote on Friday, and the customer replies that they have already ordered elsewhere. In made-to-order manufacturing, distribution, construction and technical services this is a familiar story — and quote automation targets exactly that gap between the enquiry and the sent quote, not pricing itself.

The short answer: for catalogue items, a draft quote can be ready in minutes. AI reads the enquiry from the email, matches the items to your price list, fills in prices and availability from the ERP, and the salesperson just reviews and sends it. With drawings and complex production costing it cannot do the whole job — there AI prepares the groundwork and a person sets the price.

Which route suits you depends on what your enquiries look like and where exactly the time goes.

Why quote speed matters

A customer asking you for a quote is usually asking your competitors too. Whoever replies first, and clearly, sets the benchmark — everyone else gets compared with them. And whoever replies a week later often has nobody left to be compared with.

A slow quote has less visible costs too. A salesperson who spends half the day assembling quotes is not calling customers or finding out why the last quote was lost. Enquiries that look like hard work get pushed back — and those tend to be the biggest ones.

Where time goes when preparing price quotes

The calculation itself takes minutes. The days are lost elsewhere:

• The enquiry sits in the inbox. The salesperson is in a meeting, at a customer site, or handling a complaint.
• Working out the request. The customer sends a spreadsheet with their own item names, a PDF from their system, free text or a drawing. Someone has to figure out what exactly they want.
• Searching the price list and catalogue. “Cable 3×2.5, 100 m, black” has to be found among thousands of items, the right variant chosen and stock checked.
• Discounts and terms. Which price tier does this customer get? Does a volume discount apply? Often only one person knows.
• Formatting. Copying everything into a Word or Excel template, adding lead times, payment terms, attachments.
• Approval. Larger amounts or unusual discounts wait for a manager.

Seen this way, most of the time is not spent deciding — it is spent searching and retyping. And that is precisely what can be automated.

Three ways to speed up quoting

1. Templates, a spreadsheet price list and quotes from the ERP

The cheapest step, and one many companies have never done properly: one up-to-date quote template, one price list, and quotes generated straight from the ERP or invoicing system. If salespeople currently copy prices from three spreadsheets, tidying that up alone saves a lot.

When it is enough: few enquiries, a simple range, enquiries arriving in a structured form (web form, online shop, phone).

2. CPQ software

CPQ (configure, price, quote) is quoting software for configurable products: the salesperson picks options in a form, the system applies pricing rules and generates the quote. For companies with many combinations — windows, machinery, engineered assemblies — it is a strong tool.

What CPQ does not solve is the input. Someone still has to read the enquiry email and click it into the form. Rolling it out also means encoding every rule in the system, which is often a project of several months, and licences are paid per user.

When it makes sense: configurable products, lots of rules, salespeople who live in the CRM.

3. AI on incoming enquiries

AI handles exactly the part that neither templates nor CPQ do — reading the enquiry and searching the catalogue:

1. Reads the enquiry — the email body, a spreadsheet of items, a PDF, even a scan. It extracts items, quantities, deadline and special requirements.
2. Matches items to your catalogue and price list. The customer’s “cable 3×2.5 black” is mapped to your code; where it is ambiguous, it offers two or three candidates.
3. Fills in prices, discounts and availability from the ERP according to the customer’s price tier.
4. Drafts the quote in your template.
5. Flags anything unclear. Unknown item, non-standard modification, quantity outside the usual range — none of it is resolved by guessing.

The salesperson reviews the quote, adjusts it if needed, and sends it. Nobody decides for them — they just stop searching and retyping. We describe how we design this, including the ERP connection, on the page quote automation.

When it makes sense: dozens of enquiries a week, customers who all write differently, a range with a catalogue and a price list.

These routes are not mutually exclusive. AI can break down the enquiry and pass the result into an existing CPQ or ERP instead of its own template.

Where the line is: drawings and production costing

Let us be specific. If a customer sends a part drawing and asks for a price on 500 pieces, the price depends on the process, machine time, material, subcontracting and how busy the shop floor is. AI cannot reliably calculate that and should not pretend to.

What it can do: pull material, dimensions, quantities, surface finish and deadline from the drawing and the accompanying text, find similar past jobs and what they were priced at, and hand the estimator a prepared brief. The estimator then assesses rather than retypes. A construction bill of quantities in a spreadsheet works similarly — matching items to the price list is doable, but pricing the work on a specific site stays with the estimator.

What needs to be in place

Without these, quote automation cannot be done properly by any route:

• One up-to-date price list. If three versions exist, AI will not decide for you which one is valid.
• A catalogue with codes, names and alternatives. The better the descriptions and synonyms, the more accurate the matching.
• Discount rules written down, not kept in a salesperson’s head — price tiers, volume discounts, who can approve what.
• A sample of 100–200 past enquiries together with the quotes sent in response. That is what you measure the drafts against.
• A person who reviews the quotes and has room for it in the working day.

If the price list, catalogue and stock live in different systems, it is worth connecting them first. Otherwise AI will just produce quotes from the wrong numbers faster.

When it does not pay off

• A few enquiries a week. Tidying up the template and price list costs less than any project.
• Every quote is a project of its own — one-off machines, bespoke builds. AI helps with the groundwork, but the time saved will be small.
• Prices are always negotiated individually and the price list is only a guide. You can automate reading the enquiry, not the price.
• Enquiries already arrive structured, for example through an online shop or a B2B portal. A simple rule or a quote straight from the system is enough.

What it costs and when it pays back

Indicatively, excluding VAT, for a small or medium company: analysis of enquiries and the price list €900–1,800, pilot on your real enquiries €3,500–7,000, production rollout with ERP and catalogue integration €6,000–15,000, operations from €390 a month plus model usage. The ranges are broken down in how much an AI solution costs.

A modelled example: with six enquiries a day and 45 minutes per quote, that is roughly 90 hours of work a month. If reviewing a draft takes ten minutes, around 70 hours are freed up. The bigger gain usually lies elsewhere, though — in enquiries that today are not answered in time, and in customers who do not buy from a competitor in the meantime. That cannot be calculated in advance, but it can be measured from the first day of the pilot.

Summary

• Few enquiries and a simple range: a template and a single price list.
• Configurable products with many rules: CPQ software — but input from email stays manual.
• Dozens of enquiries in different formats: AI on incoming enquiries that drafts the quote for a salesperson to send.
• Drawings and production costing: AI prepares the groundwork, a person sets the price.

The same principle works for orders arriving by email — it often makes sense to tackle both at once.

How many enquiries a week wait more than a day for a quote at your company? If you can even roughly estimate it, let us go through it together — and if automation does not pay off, we will tell you.