item_group_id errors

What Google requires

One group, varying on real variant attributes

Rows that are variants of one product share an item_group_id and differ on at least one variant attribute: colour, size, material, pattern. Within a group, identity attributes agree, the same brand throughout, and each variant keeps its own id, its own GTIN, and ideally a URL that lands on that variant preselected.

What happens if you do not

Groups that collapse or collide

Two rows in a group with identical variant attributes are duplicates of one variant, and they compete. A group whose rows never differ on any variant attribute is not a group at all. Mixed brands inside one group read as different products chained together. And one URL shared by every variant means Google cannot land a shopper on the size they searched for, which flattens the very distinction the group exists to express.

How to fix it by hand

Group by parent, differ by variant

• Use the parent product's identifier as item_group_id and give every variant row its own id and identifiers.
• Make sure the rows in a group actually differ on the variant attribute they claim, one colour value per row, one size per row.
• Link each variant to a URL that preselects it, most shop systems can, and keep brand and category identical within the group.

Which of our checks apply

5 checks apply, read from the scanner's registry

The scan evaluates groups as wholes across the file, which is how these faults become visible at all: no single row is wrong, the set is.

• item_group_id_missingat risk
• item_group_duplicate_variantat risk
• item_group_mixed_brandsat risk
• item_group_no_variant_attributeat risk
• item_group_shared_urlat risk

The quicker way

Let the scan find them

One scan reads every variant group in the file as a set and reports the groups Google will merge, split or misread. The scan to identify errors is free and shows every finding before you pay anything. You only pay for the fix.

Scan your feed