OrderSpec · Case studyOrderSpec: built from the order pages of real makers
Before writing any code, we studied how 19 Shopify makers collect engraving, embroidery and name details. This is what we found, and how it shaped an app that catches mistakes before they reach the workbench.
- Maker stores studied
- 19
- Product options app supported
- Any
- Changes made to your orders
- 0
Personalised products are good business. A name on a necklace or a logo on a hoodie sells for more than the plain version and gives customers a reason to choose a small maker over a marketplace. But custom orders go wrong in ways plain orders never do, and every mistake costs the materials, the second postage and often a refund.
We wanted to know exactly where those mistakes come from before building anything.
What we looked at
We picked 19 Shopify stores that sell made-to-order items, from engraving and embroidery to name gifts, wedding pieces and signs. For each one we went through the product page as a customer would: what the store asks for, which app collects it and what the customer is free to type.
We also read what makers say about custom orders in the Shopify Community and in reviews of the personalisation apps they use.
What we found
The details live in many different places. Some stores used Infinite Options, some Globo, Zakeke or Bold, and some a plain form in the theme. Each saves the customer’s text in a slightly different way, so anything that only works with one options app is no use to most makers.
Nothing checks what the customer typed. The options apps collect text. They rarely check it against what the maker can actually produce. A name that’s too long for the pendant, an emoji the engraving font doesn’t have or a required field left blank all go straight through to the order.
The real work happens one order at a time. Makers open each order, read the personalisation, then copy it to a list or a sheet. With forty orders that’s an hour gone before anything gets made, and every copy is another chance to misread something.
The hardest mistakes are human. “Jonh” instead of “John”. Two names typed into one field. A note saying “please make it gold, not silver” hidden in the gift message. Simple rules can’t catch these.
How that shaped OrderSpec
| What we found | What OrderSpec does |
|---|---|
| Details spread across many apps | Reads every field the customer filled in, whichever app collected it |
| No checks on what was typed | Rules for length, blank fields, fonts and allowed options, suggested from past orders |
| One order at a time | Every open custom order and its details on one screen, plus run sheets and packing slips |
| Human mistakes | AI looks at the unclear ones: likely typos, two names in one field, hidden instructions |
Three decisions came straight from the research.
- Rules first, AI second. Most problems can be caught by a simple rule, instantly and for free. AI only looks at the orders the rules can’t judge. That’s why the free plan has unlimited simple checks.
- Flag, never change. OrderSpec points problems out and never edits, cancels or holds an order in Shopify. The maker always decides.
- Only the data it needs. OrderSpec reads the product, the personalisation and the order number. It never reads customer names, emails or addresses.
Where it is now
OrderSpec is in Shopify’s App Store review. Makers who join early access get Pro free for their first three months, and we’ll publish their results here once their first busy season is through.
Sell personalised products on Shopify? Join early access and we’ll set you up the day OrderSpec goes live.