A product that comes in multiple sizes is one product with variants, not several products. In the Hyperzod CSV structure, variants live inside option groups on the product's row. Getting variants right matters because they multiply: a t-shirt in four sizes and three colors is twelve variants, twelve price points and twelve inventory records — but still one product with one SKU. This guide explains how variants are represented and how to create them with CSV Pilot instead of typing the structure by hand.
One product, many choices
Take a pizza that comes in Small, Medium and Large. The catalog should contain:
- One product row — "Margherita Pizza" with a base price (say ₹299 for Small)
- One option group — "Size" with three variants: Small (+₹0), Medium (+₹100), Large (+₹200)
A customer ordering a Large pays ₹499 — the base plus the Large delta. The store benefits too: changing the base price to ₹349 updates every size automatically, because the deltas stay the same. This is why the delta model exists — it makes the common case (adjust the base price) safe and the rare case (change one size's premium) explicit.
The variant cell format
Inside the option group, each variant is a comma-separated record:
name, price, cost, inventory, description, image
Real example for the pizza:
Small,0.00,0.00,50,10 inch,https://img/small.jpg
Medium,100.00,0.00,50,12 inch,https://img/medium.jpg
Large,200.00,0.00,50,14 inch,https://img/large.jpg
If a variant description contains a comma, the whole field must be quoted — this is the most common source of broken variant imports. A description like "10 inch, thin crust" without quotes shifts the image URL into the wrong field.
Variant rules to remember
- Deltas, not totals — the variant price is the difference over the base, never the full price.
- No variant SKUs — the Hyperzod CSV has a product-level SKU only. Variants have no SKU field.
- One group per dimension — sizes go in a Size group, colors in a Color group. Do not merge them into one group with combined values like "Black Large".
- Keep the base at the default — the cheapest or most common variant should be the base so deltas stay positive. Conversions use the cheapest variant as the base automatically.
- Required vs optional — a size group is usually required (the customer must pick one); a topping group is optional.
- Two levels confirmed, three warns — nested option groups work up to two confirmed levels. A third nesting level is allowed but flagged with a validation warning.
- Inventory in the cell is optional — a variant inventory count is only written when it is greater than zero; empty keeps the per-variant quantity stepper off.
Creating variants with CSV Pilot
Describing the variants in chat is far easier than writing VARIANTS cells:
"Create a t-shirt with sizes S, M, L and colors Black and White. Base price ₹499, Large is ₹150 extra."
CSV Pilot builds two option groups — Size and Color — with the correct deltas, quotes any descriptions safely, and shows the result in the preview. You can click any variant to adjust its name, price or image before exporting.
When converting from Shopify, the same logic applies automatically: Shopify's Option1 Name / Option1 Value columns become option groups, and each variant row's price becomes a delta over the cheapest variant. The customer-facing prices stay identical — only the encoding changes. Shopify's placeholder Option1 Value = "Default Title" rows are dropped, so simple products stay flat instead of gaining a fake option group.
The rule of thumb: if the customer must choose it before buying, it is a variant — and CSV Pilot structures it as an option group for you.