When you import a spreadsheet, any column whose heading matches one of your custom fields (by its label, ignoring capitals and punctuation) fills that field on every imported item: “Colour”, “Sub category”, “Season”, whatever you named it. The library fields (SKU, supplier, cost price, expiry…) are matched as before. A column called Category fills a custom field named Category if you have one, and still matches a tag of the same name. Create the fields first, then import, and the review screen shows the values each row will carry.
Photos too: a column headed Image, Image URL, Photo or Picture can hold one or more public https links to pictures (separated by commas). After the rows are saved, our server fetches each picture, optimises it and stores our own copy on the item (up to 8 per item), exactly as if you had added it in the item editor, so it shows in the list, in the phone apps and in Square. The import itself runs at full speed. Most pictures appear within the hour; very large catalogues can take up to a day. Links that cannot be read are skipped; the item itself is still imported. Photo links are unlisted: anyone holding the exact link can view a picture, but the links are not listed anywhere and cannot be guessed. Removing a photo takes it off the item, the phone apps, the API and Square right away and the stored file is deleted; copies on our delivery network can remain reachable by the exact link for a limited time afterwards, at most a year.
Already have the items and only want to add their pictures? Import a sheet with the names, barcodes or SKUs and the picture links, and tick “Add photos to matching items” on the review screen: each matching item that has no photo yet receives the links, and nothing else on it changes.