In a Polish SME that manufactures to order — gates, custom furniture, control cabinets, boilers, fences, steel structures — the first customer contact usually looks like this: a phone call, factory noise in the background, a salesperson scribbles dimensions on a sticky note and promises a quote "by tomorrow". Then that quote travels through a designer, a technologist and accounting. Four days. In the meantime the customer gets three other offers and picks the one that arrived first.
A B2B product configurator is the answer to that problem — but only if you understand what it actually is. Because on one hand it's one of the most obvious tools to build in a manufacturing company. On the other, it's one of the most common projects that end up as three demo screens and the words "we'll come back to this when the shop floor stops burning". This post is about how not to end up in that second group.
When a configurator makes sense (and when it doesn't yet)
A configurator pays off when you sell a variant product with a finite set of parameters. Not "a product tailored to the customer", but "a product with adjustable parameters". The difference is subtle, but crucial. An industrial gate usually has 6–10 parameters (width, height, panel material, drive, RAL colour, controls, wicket door, insulation). Each has a finite list of values. A configurator works. A complex bespoke fit-out for a specific dental clinic — it doesn't work, because there are too many exceptions.
Signals that this will pay back at your place:
- Sales writes quotes 3–15 times a day, each one at least 30 minutes of calculation.
- Enquiries repeat — 80% of quotes cover 3–4 product families.
- The designer is a bottleneck — the same person confirms the price and creates the production spec.
- Customers ask "how much?" over email and leave before anyone gets back to them.
- Discounts get handed out "by feel", because in the rush nobody calculates margin against the manufacturing cost.
If you make 5 things a day and each is different — a configurator isn't your problem yet. If you make 50 things a day and each is a variant of the same idea — it is.
What "a configurator" actually means in practice
A configurator in an SME is usually three layers that together make one tool.
Layer 1 — the interface. A form (on your website, in a B2B portal or the sales intranet) where the customer or salesperson picks parameters from dropdowns, sliders and text fields. Crucial: the system prevents impossible combinations ("3-phase electric drive" is disallowed with "2 m width" because the motor is too powerful).
Layer 2 — the rules and pricing engine. A set of rules like "if width > 4 m, the panel automatically switches to reinforced", "RAL 1023 adds 12%", "Somfy drive in variant XY adds 850 PLN". Plus formulas computing the manufacturing cost from the current material prices and standards. This is the hardest part, because that's where the technologist's knowledge lives.
Layer 3 — integration with ERP and production. The configuration result lands as a ready order in ERP (Comarch Optima, Subiekt GT, enova365, Microsoft Dynamics AX) with a broken-out BOM and routing. Without this the configurator is just a fancier Excel — the price drops from the form into an email, and production gets a piece of paper anyway.
Miss any of these layers and you don't have a configurator. You're halfway there. And halfway in this project costs more than the whole thing, because you have to maintain it by hand forever.
A real example: a 24-person industrial gate manufacturer
At one of our clients — a 24-person maker of segmented gates and industrial shutters near Wrocław, running on Subiekt GT — the quote process looked like this: 4 salespeople, each with their own price Excel in a "from last quarter, I think" version, an average of 8 enquiries per salesperson per day, time from enquiry to sent quote averaging 3 days. Enquiry close rate: 18%. The designer spent around 12 hours a week confirming feasibility of variants salespeople had invented "by feel".
We built them a configurator (web form + rules engine in .NET + integration with Subiekt through order and BOM creation). Parameter scope: width, height, panel type (5 variants), insulation (3 levels), drive (5 variants), controls (7 options), colour (RAL palette split into standard/surcharge), wicket door (yes/no/hinge type), fittings. Business rules — 68 items, collected from the designer in 3 workshops of 4 hours each.
Effect after 5 months:
- Time to prepare a quote fell from an average of 3 days to 12 minutes in 82% of cases (the rest are non-standard orders that still go through the designer).
- The designer got back around 10 hours a week — he now works on genuinely hard orders and product optimisation.
- Enquiry close rate rose from 18% to 27% — not because the offer was better, but because it arrived first.
- The portal today has around 130 active distribution partners who configure gates for their own end customers themselves. Orders land in Subiekt ready for production, with a BOM and a planned delivery date.
Implementation cost — less than the annual cost of one additional salesperson the company would otherwise have hired.
Where this kind of project usually falls apart
We've seen the same pattern several times — the project starts well, then gets stuck. Honest list of pitfalls:
- "Let's collect all the variants before we start". You never will. A configurator always gets built for 80% of variants, the other 20% keeps going the old way for now. If someone won't accept that, they'll be back in Excel a year later.
- The business rules live in one person's head. If the technologist refuses to write them down ("well, it depends"), you don't yet have a configurator — you have a project. That rule-writing workshop is valuable independently of any software.
- The material price list changes faster than configurator updates. The configurator has to pull prices from ERP (or at least from one Google Sheet that gets updated regularly), not have values hard-coded.
- No fallback for non-standard enquiries. They will always come. The configurator should have a "this is non-standard, contact us" button — otherwise customers assume you can't make a 3.7 m wide gate.
- Nobody measures how many quotes go through the configurator and how many go around it. Without that metric, after six months nobody knows whether it paid off.
Where to start
If this text reads like your company, don't start with "how much does a configurator cost". Start with three things — how many quotes you write per week, how many products they cover, and how many of them fit within 5–10 parameters. If after that analysis you see that 70% of quoting can close in a repeatable pattern — you have a project that will pay off. If everything is exceptional — you don't.
If you want to work it through honestly with someone who runs projects like this at Polish manufacturers a few times a year — the first half hour is free.
A configurator doesn't sell the product. It sells response speed. And in B2B, the first sensible offer wins two out of every three enquiries.