My print on demand size range was wrong on my own store, and the AI drafting my caption believed it. In the daily social batch for September 20, 2026, the pipeline for Marie Gems wrote that the Teaching Is the Heart of Christmas sweatshirt came in small up to 3XL, with 5XL in sport grey. Both colors actually go up to 5XL. If you sell print on demand, the size range on your store is not proof of the size range your supplier can make.
What did my store say about the sweatshirt sizes?
The live Shopify product only had Sand variants through 3XL. The batch drafted its caption from what the store showed, which is normally the safe way to do it: read the live product and trust it. Here the data was narrower than reality, so the caption came out confident and wrong.
Why was my print on demand size range wrong on the store?
The sweatshirt is a print on demand product synced from my supplier, and the blank can be made in more sizes than were switched on. The missing pieces were the Sand 4XL and 5XL variants. A store's variant list tells you what is for sale, not what is possible.
Why can I not just add the missing sizes in Shopify?
Because a variant I create by hand in Shopify has no link to the supplier. It would look real on the store, but an order for it would not be fulfilled. On a synced product, the supplier decides what exists, and the store only mirrors it.
How can you tell a product is synced from a supplier?
In my store, a print on demand product carries a link to the supplier's version of it, and its inventory is held at a supplier location. That is the signal to stop and check the supplier before changing anything.
Where does the fix actually go?
Through the supplier: enable Sand 4XL and 5XL there, then sync them to the store. Once those variants are live, the sweatshirt's captions go back to saying small up to 5XL.
What rule does my social pipeline follow now?
When a caption or slide has to state a size or color range for a print on demand product, the pipeline checks the variant list and says plainly when it looks narrower than the blank supports, instead of writing the caption around the gap. It also never creates variants directly in Shopify on a supplier-managed product. The rule is written into my lessons file, so the next draft starts from it.
What should a caption say while the size list is being fixed?
It should say plainly that the variant list looks narrower than the blank supports, instead of writing the caption around the gap. A caption that quietly writes around a gap looks fine and leaves the real problem in place.
Why did an AI believe my store?
It did what I would expect a careful assistant to do: it read the only source it had. That is the part I have to design around. An AI drafts from the data it is given, so wrong data becomes a confident caption. The fix is not asking it to be more careful. It is a written rule that tells it when a number needs a second source.
What should another seller check before writing a size range?
Compare the variants on your store with what your supplier lists for that blank, and treat any difference as an open question, not a fact. Fix it at the supplier and let it sync. For where AI fits in my own work, what I actually let AI handle in my businesses draws the line, and how I use AI agents across my businesses shows the pipelines it sits inside. My Brands page covers what they support.
Frequently asked questions
Why can I not add a missing size directly in Shopify on a print on demand product?
A variant created by hand in Shopify has no link to the supplier's product, so an order for it would not be fulfilled. On a synced print on demand product, the missing size has to be enabled at the supplier and synced to the store.
Is my store's variant list proof of the sizes a product can be made in?
No. The list only shows the variants that have been switched on. My Teaching Is the Heart of Christmas sweatshirt showed Sand variants only through 3XL, while both colors can be made up to 5XL.