AI shopping product data checklist
Review six areas: identity, content, taxonomy, variants, assets and optional enrichment. These are Product Feed Scan coverage dimensions, not a universal AI platform certification.
Fix foundational data first
Start with stable unique product IDs, factual titles and descriptions, current commercial data and usable image references. Resolve missing or malformed fields before optional enrichment. A detailed description cannot compensate for a missing item identifier or an invalid price.
Review applicability
A product without variants should not be forced into a variant group. Add material, color, size or extra images only when they accurately describe the item. Empty or invented values can satisfy superficial formatting expectations while making the catalog less trustworthy.
Use dimensions as a work queue
Open the dimension with the largest observed coverage gap, review affected fields and repair the source mapping. Then regenerate and rescan. Keep a before-and-after result using the same ruleset. A higher indicator shows improvement against these checks; it does not prove increased AI visibility.
Practical example
Synthetic checklist: a plain mug has a stable ID and image but no category. Add the correct category from catalog data. Do not invent a size variation or extra image just to increase an enrichment indicator.
What Product Feed Scan checks
The general audit computes deterministic identity, content, taxonomy, variant, asset and enrichment coverage. Text length and field presence are proxies; semantic truth, live image quality and recommendation outcomes are not measured.
Open the relevant checker See a sample report