AI sizing that speaks Shopify.
Garu works from the catalogue you already have - your variants, your size labels, your charts - rather than asking you to rebuild them.
The catalogue work is the part that usually kills these projects.
Most sizing tools fail on Shopify for an unglamorous reason: they need clean, structured garment measurements that almost no store actually has. The pilot stalls at a spreadsheet, someone is asked to key in chest and waist figures for four hundred SKUs, and the project quietly dies.
Garu starts from what is already in your store. It reads your product variants and the size labels you actually sell - S/M/L, US 4, W32, EU 38 - and works against the size chart you already publish, wherever it lives. Where a collection uses its own chart, that mapping is respected instead of flattened into a store default.
The practical effect is that enabling sizing on a new collection is a toggle, not a data project. You can start with one collection, see whether the recommendations hold up against your own returns data, and widen from there.
- Your variants
- Recommendations come back in the size labels you actually sell.
- Your charts
- Per-collection charts are honoured, not flattened to a store-wide default.
- No re-tagging
- No measurement spreadsheet required before you can switch it on.
- Roll out gradually
- Enable one collection, check it against your returns data, then widen.