Name your category pages after what shoppers want, not what you sell. A page called "jeans" matches a product; a page called "airport outfits" matches an intent, and intent is how people, and AI assistants, actually ask. As shopping moves into conversational AI, category and navigation pages built around shopper goals and situations are far easier to retrieve, cite, and recommend than pages named after your internal catalogue taxonomy. This is one of the highest-leverage structural moves an ecommerce site can make for AI visibility, and it sits above the product page, at the level of how your whole store is organised.
The pattern was a recurring theme in a widely-shared BrightonSEO April 2026 recap published by Peec AI, an AI-visibility analytics vendor. The reported results are single-vendor, conference-sourced figures, so treat the numbers as directional, but the underlying mechanism holds up against how query fan-out works.
Why do product-named category pages underperform in AI shopping?
Product-named pages underperform because they answer the taxonomy question, not the shopper's question, and AI assistants search on the shopper's question.
When someone asks an assistant for help, the engine expands the request into many natural-language sub-queries, a process of query fan-out. Those sub-queries are phrased the way people think ("what should I wear for a long flight?", "trail shoes for wide feet"), not the way your merchandising team files inventory ("bottoms > denim > slim"). A category page that matches the intent language has a clean anchor to be retrieved against; a page named only for the product type does not.
Two shifts make this sharper in 2026:
- Shoppers ask, they don't just search. Conversational interfaces invite full-sentence, situational requests, so the gap between customer language and catalogue language is now exposed on every query.
- The long tail is getting longer. As assistants handle more of the phrasing, the head terms shrink and the specific, situational queries multiply, exactly the demand that intent pages capture and product-type pages miss.
What does an intent-driven category page look like?
An intent-driven page is organised around a goal or situation and assembles products from across your catalogue to serve it. The contrast is concrete:
| Product-named (taxonomy) | Intent-driven (shopper goal) |
|---|---|
| Jeans | Airport outfits for long-haul comfort |
| Sneakers | Road shoes for wide feet |
| Dresses | Wedding-guest outfits under a budget |
| Backpacks | Cabin bags that fit carry-on limits |
| Blenders | Blenders for smoothies with frozen fruit |
Each intent page pulls the right items together, explains who and what it is for in plain language, and answers the natural follow-ups (fit, occasion, constraints) on the page. The BrightonSEO recap described a large fast-fashion retailer that moved from traditional product categories to intent-driven ones and reported a very large traffic uplift, with the featured items selling out. Read the specific percentage as directional single-source data; the reproducible lesson is that matching the page to the intent, not the product noun, is what unlocked the demand.
How do you use the customer's own words in navigation?
Use the words your customers use, not the expert terminology your team uses internally, because the label is what both humans and AI match against.
The same recap reported that changing navigation from internal expert terms to customer language lifted key events and revenue for one retailer. The figures are single-source and should be hedged, but the principle is durable and testable: internal jargon ("technical outerwear", "occasion footwear") is invisible to a shopper who asks for "a warm waterproof jacket" or "comfy shoes for standing all day". Steps that work:
- Analyse problems, not keywords. Start from the shopper's pain point or situation, then find the language they use to describe it.
- Mine the fan-out. List the natural-language sub-questions your category triggers, and look for the common terms and concepts that recur, per the content loop for fan-out queries. Those recurring phrases are your page titles and navigation labels.
- Rename to match. Relabel navigation and category titles in that customer vocabulary, keeping the product-type structure underneath for classic search and site plumbing.
- Answer the branches on the page. Cover the situational follow-ups (fit, use case, constraints) so the page is self-sufficient when an assistant lifts from it, the same discipline as writing product pages AI can quote.
Your customers don't shop by your org chart. Name the page after the need they typed, assemble the products that meet it, and you have built the exact thing an AI assistant is trying to retrieve.
How does this fit with the rest of your AI shopping work?
Intent-driven category pages sit above your product pages and your feed, and they complement both rather than replace them.
- Feed and catalogue enrichment (preparing your catalog for AI agents) makes each product understandable to an agent.
- Conversational product pages (writing PDPs AI can quote) make the individual item liftable.
- Intent-driven category pages make the collection discoverable for the situational, natural-language queries that assistants fan out into, and connect to how Google AI Mode shoppers ask in attributes.
Done together, they cover the fan-out at every level: the situation, the comparison, and the specific product. Start with a handful of intent pages for your highest-value shopper situations, measure whether AI assistants begin surfacing them, and expand from what works.
Knowing whether these pages actually get cited and recommended, rather than just built, is the hard part. That is where Buffy Intel fits: it snapshots whether AI shopping assistants surface your store for the situational queries your customers ask, so you can see which intent pages earn recommendations and double down on the ones that do.