In most SMEs that call us in because "sales can't keep up", the bottleneck isn't sales. It's that to send a quote covering 30 line items, a salesperson needs two hours, a calculator, three Excel tabs and a quiet prayer that the price list hasn't changed since Monday.
This text is about why quoting is so often the bottleneck in small and mid-sized Polish manufacturers and distributors — and what you can actually do about it without buying a 200,000 zloty CRM. The view from someone who builds these projects every few months, not from someone who sells them on slides.
How to tell that quoting is eating your company
Usually no one inside the company says "we have a quoting problem". They say: "we miss customer deadlines", "we lose enquiries we didn't get to this week", "the sales rep never has time to visit clients". All of those sentences hide the same root cause. The signs we see most often:
- A salesperson issues 1–3 quotes a day and each one takes them at least an hour.
- Everyone has their own Word or Excel template — and each one is "slightly different".
- The price list lives in one file, "discretionary" discounts live in the owner's head, and the margin only surfaces in the bookkeeping.
- After a quote goes out, no one knows whether the customer opened it, where they got stuck, or when to follow up.
- Half of the orders that hit production have some discrepancy versus the quote — because someone retyped something along the way.
Each of these signs looks innocent on its own. Together they mean your company is losing orders not because it's expensive or slow — but because the quoting process is the slowest link in the sales chain.
What's actually eating those 2 hours
When we look closely at how a typical quote is born in a 20–50 person manufacturer, it's made of several steps almost none of which require "thinking" — and yet each one eats fifteen minutes.
- Retyping the customer's enquiry from the email into a template (5–10 minutes).
- Finding current prices for products or components. Usually in three places: ERP, an Excel discount sheet, an email from the supplier with the latest change (15–30 minutes).
- Calculating margin, roundings, transport, sometimes customs or packaging (15–20 minutes).
- Checking with the boss whether a discount is allowed (sometimes a day, if the boss is out).
- Formatting the table so it doesn't collapse in PDF (10–20 minutes, depending on the phase of the moon).
- Sending the email, saving a copy in the customer folder, adding a row to the "quotes 2026" spreadsheet.
Total: 60–120 minutes. Of which the actual intellectual work — "what should we really propose to this customer" — is often 10 minutes. The rest is logistics. That rest is exactly what should be automated first.
Real example — from 2 hours to 10 minutes
For one of our clients — a 22-person PVC window manufacturer — issuing a quote took on average 90–120 minutes. The customer (a developer or a contractor) sent an enquiry by email, sometimes with a technical drawing, sometimes just a table of dimensions. The salesperson manually priced each window by profile, additional rails, glazing and fittings, checked prices with two glass suppliers, retyped everything into Excel, dropped in a logo, saved a PDF, sent it. Any change from the customer — "what if we drop from triple to double glazing?" — added another half hour.
We built them a simple configurator in .NET — a form where the salesperson enters dimensions, picks profile, glazing, colour and fittings. The system pulls current prices from Comarch Optima through its API, applies the margin from a rule written once in an admin panel, generates a PDF in a uniform format and stores the quote in a database. The salesperson can override a price or add a discount manually, but in 80% of cases the quote goes out in 7–10 minutes, with no effort.
The result after six months: average quote time dropped from 105 to 12 minutes. The number of quotes sent per week rose from 18 to 47. Quote-to-order conversion went up slightly (from 22% to 27%) — mostly because customers got a reply the same day instead of three days later. The project paid for itself in under 5 months.
This isn't a spectacular case. It's the typical effect when you take a process where 80% of the work is shuffling numbers between windows and replace it with a form plus rules.
How it works in practice — without "AI" in the name
We like it when a conversation about quote automation doesn't start with the word "AI". In 90% of cases you don't need any model — you need deterministic logic that someone has to write down once. Price = (component cost × quantity) + margin per customer group + transport from a table + discount if the order threshold is met.
Such a solution typically has four parts:
- Form or configurator. Internal (for the salesperson) or external (for the customer on the website). In SMEs we usually start with the internal one — faster, cheaper, less risk.
- Pricing and discount rules. Pulled out of the owner's head and written into a database. This is the hardest part of the project because it forces decisions that until now were left hanging ("well, depends who's asking").
- ERP integration. Comarch Optima, Subiekt GT, enova365, Microsoft Dynamics AX — they all expose APIs or readable databases. Without this the prices will diverge within two weeks.
- PDF generator and quote registry. So the salesperson doesn't burn time on formatting and so someone can later check what was sent to whom.
This is not a year-long project. A first version that genuinely takes time off the sales team is usually 6–10 weeks and a budget in the tens of thousands of zlotys. Less than the annual cost of one extra salesperson you'd otherwise have to hire to handle the same volume of enquiries.
What not to automate
Not everything that can be automated is worth automating. A few traps companies fall into when they think "let's put a configurator for customers on the website":
- Negotiation. If 30% of your quotes end in a phone call about price, a customer-facing configurator changes very little. Speed up the internal process first before exposing anything to the outside.
- Non-standard pricing. If every project is genuinely different (e.g. steel structures built to bespoke specs), a configurator will cover 50% of cases and the rest will still go through manually. That's still a gain — but don't expect 100%.
- Strategic decisions. The margin for a new strategic customer should not be calculated by a system. That's an owner's call, not a rule.
- Processes nobody can describe. If three people in the company give three different answers to "how do we price this?", agree the answer first. Automating chaos gives you chaos at higher resolution.
Where to start
Don't start with "let's build a quoting platform". Start with one number: how long does it take you to issue an average quote today, and how many quotes do you issue per week. Multiply by the salesperson's hourly cost. That's your annual cost of "manual quoting" — usually a number that surprises the owner.
Concretely:
- Measure: how long a typical quote takes and the annual cost of the person who issues it.
- Map: where the input data comes from (email, phone, EDI), where the prices live, who approves discounts.
- Pick one segment of quotes (e.g. "typical quotes for catalog products") and work out what percentage of all enquiries it represents.
- Start with a configurator for the salesperson, not for the customer. Get the internal process right first.
- After go-live, measure again. Without before-and-after numbers, no one in the company will believe the project delivered anything.
If you don't know what manual quoting really costs you today or what's worth automating first, we'll help you map it. The first call is free and takes half an hour.
A quote sent three days late is a lost quote. In quoting, speed beats price more often than salespeople care to admit.