ChatGPT understands your products better than you do – and here’s why that matters

Built for machines, not for people

The standard eCommerce catalogue is built around specifications: colour, size, material, price. All of it is factual, technical and easy to ingest into Google Shopping or a supplier feed. But that’s not how many people shop.

Nobody starts by thinking; "I need a linen blend midi dress with a side zip." They search for something to wear to a summer wedding in Tuscany. They don’t want a "rubber-soled leather loafer" - they want something that works for the office and date night. This is the gap between how most catalogues are structured and how customers actually think.

That gap is getting wider

With the rise of AI search, that disconnect is only becoming more costly. These systems can reason about use-case. They understand that "lightweight + sleeveless + linen" implies "holiday-ready". They know that "machine washable + cotton + pastel" might signal "nursery friendly".

But if your catalogue - and subsequently your site’s front end -  doesn’t surface these ideas, your site won't get surfaced either. These AI systems depend on structured product data. If it isn’t there, how will new customers find your site? 

Curation is what makes scale work

The value of a large product range is the variety of options available. But unless it's well-structured, that range can become overwhelming rather than helpful. That’s what taxonomy is for. It's what filters are for. It’s the difference between showing someone 3,000 options and showing them the three that make sense for their life.

This is especially important for Shopify stores, which don’t natively support parent-child categories. Without good structure, navigation breaks down fast. So the answer isn’t more filters or more copy. It’s smarter structuring of the data that powers everything.

How to surface emotional context on a Shopify store

1. Use-case driven collections (selectively)

Creating curated collections like “Best for weddings abroad” or “Perfect for city breaks” can help bridge the gap between what you sell and why someone buys. These pages typically outperform blog-style listicles in SEO and convert better on-site.

But from a UX perspective, you have to be careful. Shopify doesn’t allow dynamic subcategory structures out of the box, so if you create too many of these collections, navigation becomes bloated and search performance suffers.

Use static collections for:

• High-intent, high-volume use cases

• Strategic entry points to the catalogue

• Campaigns you want to scale through SEO and ads

Use filters and soft attributes for everything else.

2. Structure product data using metafields, not tags

If we were building a Shopify store from scratch today, we’d use metafields for every meaningful attribute: material, style, fit, size, use-case, occasion, care type. Tags would be used only for internal logic or campaign-based merchandising.

Why?

• Metafields give you structured, indexable, machine-readable data. They support filtering, schema markup, content generation, and collection creation - and can also be used for consistent merchandising logic when applied systematically.
  
  

• Tags are best suited to transient or ad hoc needs - like temporary campaigns, theme display logic, or quick merchandising without changing core data.

This separation creates clarity. Metafields offer structure and consistency. Tags offer flexibility.

3. Untangling legacy setups

Most established Shopify stores weren’t built this way. Before metafields were widely adopted, merchants relied on colon-based tags like style:classic or colour:navy. These served double duty – powering both taxonomy and display rules – and that approach works to a point.

But at scale, it breaks:

• Too many tags make filtering messy and unreliable

• Managing indexable collections becomes difficult

• Any theme or feed logic tied to tags becomes brittle

When we restructure large catalogues, one of the first steps is identifying where taxonomic data can be migrated into metafields. This doesn’t have to be all-or-nothing. A category-by-category migration is often the most realistic path.

Once the structure is in place, everything becomes easier:

• Collections can be built programmatically

• Shopping feeds become cleaner and more powerful

• GA4 tracking is easier to configure and segment

• AI systems like ChatGPT can surface better matches

• Internal search and recommendations become more accurate

Design patterns that bring this to life

Best-for-X labelling

In product cards, filters, and homepage modules, soft labels like "Best for summer evenings" or "Holiday ready" can bridge the emotional gap. These aren’t hard attributes, but they act as signposts for intent.

This works especially well in:

• PLPs with editorial-style headers

• Homepage slots with curated tiles

• Filter dropdowns with use-case groupings

• Badges on product carousels

Curated commerce pages

Replace the old SEO listicles (“10 best white shirts for summer”) with permanent collection pages that:

• Use structured data to curate the right products

• Include short copy explaining the selection logic

• Are designed to support both SEO and conversion

These pages behave like editorial content but are built with native Shopify logic. They’re easier to maintain and perform better across paid, organic and on-site UX.

Invest in structured data now – for what’s coming next

The benefits of surfacing emotional context aren’t limited to filters and collections. That structured data will be the foundation for the next wave of AI-driven product discovery.

We don’t yet know exactly how generative shopping experiences will evolve. But we do know what they’ll need: structured, interpretable product data that reflects how people actually shop.

Imagine a future where a customer can describe what they want in natural language – “I need something elegant but breathable to wear to a wedding in Tuscany in late May.” If your catalogue has been enriched with well-modelled use-case attributes, it becomes dramatically easier for AI tools to understand what products meet that brief.

That’s the real payoff: structured data doesn’t just support today’s UX patterns or unlock tomorrow’s discovery models - it also enables more informed UX and UI decisions right now. If your store can express not just what a product is, but why it’s right for someone, you’ll be in a far stronger position to support both human users and machine-driven discovery.

What this all adds up to

Your catalogue shouldn’t just describe what a product is. It should help people understand what it’s for. Structured properly, even a 25,000 SKU store can feel personal and navigable. But only if the taxonomy reflects real use cases, emotional context and the language customers actually use.

As AI changes how people discover and choose products, the winners won’t be the brands with the biggest ranges. They’ll be the ones who help people find what they really need, faster.

Written by Sam Wright, founder of Blink SEO - a specialist agency for large catalogue stores with a focus on Shopify SEO and Shopify PPC.