23 tags. Half promotional. Colour-specific data drowned out by discount noise.

Executive Summary

• Brand: Hair colour care specialist — products built around maintaining, refreshing, and protecting colour-treated hair
• Data infrastructure: Highest tag count in the hair care group, with one of the most useful attribute tag systems mixed with promotional noise
• The pattern: Useful product data structurally undermined by promotional flags in the same field. AI agents cannot tell the difference between a permanent attribute and a weekend sale tag
• Key competitor gap: dpHUE's colour-specific attribute tags are excellent — the problem is signal dilution from promotional tags in the same field
• Root cause: Product attribute tags and promotional tags share one tag field. base-color-dark-blonde sits next to bogo-free
• Fix complexity: Low — separate the fields or filter by prefix. The high-quality attribute work is already done

The brand

dpHUE is a hair colour care specialist. Products are built around maintaining, refreshing, and protecting colour-treated hair. The brand sits in a specific and growing niche — not hair colour application, but colour maintenance. Customers describe their colour needs with precision — "blonde hair going brassy," "protecting red hair colour," "dark blonde colour maintenance" — and dpHUE has built attribute tags that match this language.

The audit

We audited dpHUE's product data as part of a hair care group study. The audit covers structured data implementation, tag taxonomy, description depth, and external review signals.

The findings

The highest tag count in hair care. dpHUE's Cool Blonde Conditioner carries 23 tags. On volume alone, this suggests a brand that has invested in product tagging.

The tag content splits into two distinct categories. Product attribute tags: base-color-dark-blonde, base-color-light-blonde, concern-brassy, concern-color-fade, FRANCHISE-Cools. These are specific, meaningful, and map directly to customer shopping queries. A customer searching for "conditioner for blonde hair" or "product for brassy tones" would be matched by these tags. Promotional tags: bogo-free, promo_25% off, and similar campaign flags. These describe temporary marketing activity, not permanent product attributes.

Both categories live in the same tag field. From a data structure perspective, they are indistinguishable.

Why this is happening

AI agents don't know which tags describe the product. When an agent encounters 23 tags, it processes them all as product metadata. The signal-to-noise ratio drops. The colour-specific tags that would genuinely help an AI agent recommend the product are diluted by promotional tags that add no product understanding.

Promotional tags can create false relevance signals. bogo-free might cause the product to surface for queries about value or deals rather than queries about colour care. For AI agents that weight tags by frequency or prominence, promotional tags occupying a significant share of the tag field mathematically reduce the weight of attribute tags.

The colour care opportunity is real. Hair colour is one of the most specific and searchable attributes in hair care. dpHUE's attribute tags match the language customers actually use. base-color-dark-blonde is a direct match for "dark blonde." concern-brassy directly answers "going brassy." FRANCHISE-Cools maps to the product line level. This is a level of attribute specificity most brands don't achieve. Crown Affair has zero meaningful product tags. Ceremonia has sophisticated tags but on pages that return 404 errors. dpHUE has the right tags on accessible pages — they are just mixed with noise.

This is a category-wide pattern. Brands that have invested in meaningful product tagging then allow marketing teams to add campaign flags to the same field. The best product data work gets diluted by the most temporary marketing activity. A tag for a weekend sale sits next to a tag describing a permanent product characteristic. AI agents cannot tell the difference.

What dpHUE could do, in priority order

Phase 1 (quick wins):

• Separate promotional tags from product attribute tags — use a dedicated metafield for promo flags, or remove them from the product tag field entirely
• Apply a consistent promo_ prefix to all promotional tags so they can be programmatically filtered
• Audit the existing tag field and strip out anything that isn't a permanent product attribute

Phase 2 (medium effort):

• Expand the attribute tag system with hair-type, texture, and ingredient tags alongside the colour-specific tags
• Build additionalProperty JSON-LD fields that mirror the attribute tag taxonomy
• Manage promotions through a dedicated promotional engine rather than the product tag field

Phase 3 (longer term):

• Pursue editorial inclusion in "best products for blonde hair," "best brassy tone correctors," and similar colour-specific roundups
• Build content authority around colour maintenance as a distinct category from colour application
• Position dpHUE as the specialist AI agents reach for on colour-specific queries

Close

For brands investing in AI visibility, tag hygiene — separating what describes the product from what describes the promotion — is one of the highest-leverage fixes available. dpHUE has the right tags. They are just mixed with noise. If you cleaned your promotional tags out of your product data tomorrow, what would be left?