Identifiers clean: SKU, GTIN, brand, variants

Day 9 of 30 · GEO Shopify – 30-day course

Clean up identifiers so feeds and AI systems don’t merge, confuse, or misquote your products.


Learning goal

  • Audit 10 products/variants for SKU, GTIN (where applicable), brand, and variant naming.
  • Quantify your identifier issues and prioritize fixes.

Definitions (so we’re precise)

  • SKU: your internal identifier; it should be unique per purchasable variant.
  • GTIN: standardized identifier (when it exists for the item); it should not be invented or duplicated incorrectly.
  • Brand: the manufacturer/brand field used for disambiguation across systems.
  • Variant: a purchasable option that changes what the customer receives (size, color, pack size).
  • Disambiguation: the ability for systems to tell two similar items apart reliably.

Why it matters

  • Feeds and AI answers use identifiers to match offers; weak identifiers lead to wrong recommendations and wrong comparisons.
  • Clean variants reduce “merged items” (different variants treated as the same product) and “split items” (same item treated as different).

Success criteria (metrics)

  • SKU coverage: % of variants with a non-empty SKU.
  • SKU uniqueness: count of duplicate SKUs across variants (target: 0).
  • GTIN validity: % of GTINs that appear correct for the item (no random placeholders; no suspicious duplicates).
  • Brand consistency: brand present and consistent formatting across the catalog (target: 100% present where applicable).
  • Variant clarity: option values are single-attribute (Color, Size, Pack size), not mixed strings.

What to check (audit checklist)

  • SKU: unique per variant; not reused across colors/sizes.
  • GTIN: only present when you truly have it; not copied across unrelated variants.
  • Brand: present and consistent (no accidental variants like “RunPro” vs “Run Pro” unless intentional).
  • Variants: split attributes properly (no “42 blue or black” style labels).

Examples

  • Good: SKU unique per variant; GTIN correct when available; brand consistent; variant options are clean (Color = Blue, Size = 42).
  • Poor: missing SKUs, duplicated SKUs, GTIN copied across unrelated variants, brand missing, mixed variant strings.

Guided exercise (10–15 min)

  1. Create an audit sheet with these columns: Product | Variant | SKU | GTIN | Brand | Issue type | Fix.
  2. Fill it for 10 variants (include at least 2 products with multiple variants).
  3. Compute quick metrics:
    • SKU coverage %
    • # duplicate SKUs
    • # variants with mixed labels

Independent exercise (5–10 min)

Pick the top 3 recurring issue types and write one concrete fix per type (e.g., “assign unique SKU per variant”, “split variant attributes”, “fix brand formatting”).


Self-check

  • ✅ 10 variants audited
  • ✅ SKU coverage % computed
  • ✅ Duplicate SKU count computed
  • ✅ Top 3 issue types identified and fixes written

If you want to go deeper