What 15 years in the curtain trade taught me about pricing software

Most software for made-to-measure products is built by people who never ran a workshop. You can tell within an hour. The features are correct. The maths is correct. The quotes come out looking right. But the workshop cannot use them, because the software never had to survive contact with a retailer running an order at 4 pm on a Friday with a fabric that arrived short.

I ran a curtain shop before we published the Curtain and Blind Calculator as a Shopify app. Fifteen years of that shop is what our calculator has under it, and it shows up in small decisions people who never worked a shop floor would not make.

Pricing tables belong to the retailer, not the software

Every quoting tool built by people outside the trade eventually decides it should hold the master pricing. Retailers hate this. Their pricing comes from a mix of supplier deals, seasonal changes, and how they price against their local competitor. Software that assumes there is a canonical price for a fabric loses the accounts that price against their region. We keep pricing tables uploadable and changeable because that is how the retailer actually works.

Fabric width matters more than fabric price

Every shopper thinks a wider window costs more because it needs more fabric. That is true, but the step function is the number of fabric widths seamed together, not a linear price per centimetre. Software that hides this from the retailer prices small windows over and large windows under. Software written by someone who has cut fabric prices the width bands explicitly.

Metric and imperial are not a translation problem

A UK shop that takes a US order does not need a French translation of their calculator. They need a metric-to-imperial conversion that lands on the same pricing band the workshop would have hit. That means the unit toggle has to interact with the rounding rule at the band boundary, not just multiply by 2.54. This is a small thing until you get it wrong on a 60-inch order.

Setup is where retailers churn

The first three days after a retailer installs quoting software is where every install is won or lost. Not the marketing page. Not the feature list. Setup. We answer the setup call ourselves because software that hides its team behind a support portal in the first week loses every account that hit any friction in the pricing-table upload. Founder-level support in setup is not a scaling problem; it is what keeps the churn rate honest for the first hundred customers.

What all of this means

You cannot build software for a trade you never worked in. Or you can, but you will lose to the person who did. That is the honest answer to why we ended up as one of the tools this trade uses.