Executive summary
- What: the attributes applied when you categorise a product, colour, material, fit, fabric, size and the rest, are now selectable in Analytics as dimensions and filters alongside your custom metafields.
- Why: product attribute performance previously required an export. It is now answerable in Reports, Explore and the query editor.
- Who: merchandising and buying teams at merchants with large, well-categorised catalogues.
What changed
Category metafields come from Shopify's standard product taxonomy. When you assign a product category, the attributes attached to that category become available on the product, and most merchants have been populating them for discoverability on the storefront, on marketplaces and in search. The data was accurate and structured, it simply had nowhere to go in reporting.
Those attributes now appear in the Analytics field picker next to custom metafields, in their own group. You can group a sales report by Colour or Material, filter to a specific value such as a single fabric, and use the same fields in Explore and the query editor. Nothing about categorisation changes, and custom metafields behave exactly as they did. One detail matters for accuracy: when a product carries more than one value for an attribute, Analytics keeps those values together on a single row so revenue is not double counted across them.
Why it matters
Most enterprise catalogues already hold this data at a quality level nobody was using. Teams filled in taxonomy attributes because the storefront filters and the marketplace feeds needed them, then answered merchandising questions from a separate, hand-maintained tagging scheme because that was what reporting could see. Two vocabularies describing the same products, one authoritative and invisible, one visible and drifting.
Collapsing that is worth more than the individual reports. It removes a class of tag hygiene work, and it makes attribute performance comparable across a catalogue rather than across whatever a category manager happened to tag. Merchants running large catalogues with high SKU counts will feel this most, since that is exactly where manual tagging degrades fastest.
The single-row treatment of multi-value attributes is a deliberate accuracy choice and worth understanding before you present the numbers. A product available in three colours contributes its revenue once, not three times, so totals reconcile against your sales reports.
Role-specific impact
- Marketers: attribute-level demand signals become available for campaign and collection planning without an analyst in the loop.
- Developers: nothing to build. The useful contribution here is auditing categorisation coverage, since products missing a category contribute no attributes and quietly skew a report.
- Store admins: treat this as a reason to finish categorisation on the long tail of the catalogue. Reporting quality now depends directly on it.
Use-case example
Real-world scenario
A home textiles merchant carrying about 9,000 SKUs had been reporting fabric performance from a legacy tag set applied at launch and rarely maintained. Roughly 18 percent of active products had drifted out of sync with their actual taxonomy attributes, which meant the fabric mix report the buying team relied on was quietly wrong. Grouping the same sales report by the Fabric category metafield produced a different ranking, with one blend performing well above its tagged share. The team reallocated open-to-buy against the corrected view and retired the legacy tags entirely, removing a recurring maintenance task from the merchandising calendar.
Implementation checklist
- Check what share of active products carry a product category, since uncategorised products contribute no attributes.
- Identify any parallel tagging scheme built to answer the same questions and mark it for retirement.
- Rebuild your most-used attribute report in Analytics and reconcile the totals against the existing export.
- Note which attributes are multi-value on your catalogue and brief stakeholders on the single-row behaviour.
- Complete categorisation on the long tail before publishing the new reports internally.
- Set a cadence for reviewing categorisation coverage as new products are onboarded.
FAQ
Q: Do we need to turn this on or re-save our products?
A: No. If products are already categorised, the attributes they use are present in Analytics with no action required.
Q: Will a product with several values for one attribute inflate our revenue totals?
A: No. Analytics keeps multiple values for the same attribute on one row specifically so sales are not counted twice.
Resources
Category metafields in the Shopify Help Center
Need guidance? Talk to Makro.