The three jobs category pages do today
Job 1: Browsing for undecided shoppers.
Not everyone knows what they want. Many shoppers want to look around first. A category is a guided starting point.
Job 2: Landing pages for search engines.
Category pages often rank for broad, high-volume terms like “women’s running shoes.” They carry years of links and authority.
Job 3: Merchandising control.
They are where teams decide what sells first: new arrivals, margin leaders, seasonal picks.
Any forecast that ignores one of these jobs will lead to bad decisions. In particular, deleting category pages to “go AI-first” would wreck job two overnight.
AI Search Grader by bCloud AI
Grade your ecommerce search in 10 quick questions
31% of ecommerce searches return zero results — and most shoppers who hit a dead end leave for a competitor. How does your store's search stack up?
Answer 10 short questions and get your AI search score, plus a personalized report to fix the gaps. Free, takes about 2 minutes.
No signup needed to take the quiz.
Understanding intent…
Scoring your answers across relevance, AI, experience, and insights.
Your AI search score is ready
Tell us where to send your personalized report. You'll see your score and recommendations right away.
Your score by pillar
Personalized recommendations
Fix the gaps in weeks, not quarters
bCloud AI replaces keyword-only search with hybrid AI retrieval — sub-200ms responses, 99.99% uptime, and conversion lifts of up to 40% across 50+ implementations.
Why the question is being asked now
Three changes make fixed navigation look dated.
First, shoppers now type full sentences. Search has moved from keywords to intent, so a shopper can skip the tree entirely. See query understanding.
Second, retrieval can build relevant sets on demand. With hybrid keyword and vector search, “breathable hiking boots for wide feet under $150” returns a sensible collection that no merchandiser ever created.
Third, headless architectures make pages flexible. When the storefront is decoupled from the platform, a page can be assembled from a query instead of a fixed category ID.
Together, these trends explain why “ecommerce search platforms without category pages” is now a real search topic.
5 ways AI reshapes category pages by 2030
1. Ordering becomes personal
Today, most category pages sort the same way for everyone. Soon, the order will reflect who is looking.
A returning shopper who buys mid-price trail gear should see those products first. A first-time visitor should see broad bestsellers. Same category, different first row. See vector search personalization.
2. Collections are generated from intent
Fixed categories will be joined by generated ones. A popular search like “gifts for new hikers” can become a landing collection automatically, built from products that match the intent.
This is where AI browsing tools come in. bCloud’s AI Browse and Generative Experiences are built for exactly this kind of intent-driven page.
3. Filters adapt to the category and the query
Static filter lists show the same facets regardless of context. Adaptive filters show what matters for this set of products and this shopper’s query. For boots, width and waterproofing. For tents, capacity and weight. See faceted search.
4. Conversation sits on top of browsing
Browsing and asking will merge. A shopper on a category page will be able to type “which of these are best for wet trails?” and have the page re-rank in place. See conversational search engine.
5. The taxonomy moves behind the scenes
Here is the counterintuitive part. Taxonomy becomes more important, not less. It simply becomes less visible.
AI retrieval, recommendations and assistants all lean on clean category relationships to understand what a product is. So the tree stops being the shopper’s main path, yet remains the data structure everything else depends on. That is a product data job as much as a design job.
One category, three shoppers
An example makes the shift easier to picture. Take a single “Hiking Boots” category and three visitors.
Shopper A
arrives from Google, undecided. Today, they see 140 boots sorted by popularity. Tomorrow, they see a short guide at the top, filters for terrain and width, and bestsellers across price tiers.
Shopper B
is a returning customer who always buys wide-fit, mid-price gear. Today, they see the same page as Shopper A. Tomorrow, wide-fit options in their price range rise to the first row.
Shopper C
types a question on the page: “Which of these are best for wet trails in spring?” Today, nothing happens. Tomorrow, the set re-ranks in place and highlights waterproof models.
Notice what stayed the same. Same URL, same category, same merchandising rules. Only the experience on top changed.
Category pages in B2B catalogs
B2B stores feel this shift differently. Their category trees are often deep and technical. A buyer may need to click five levels down to reach “stainless steel compression fittings, half inch.”
Consequently, many professional buyers skip navigation and search by part number or specification instead. For them, the tree mostly matters as data, because it tells search and recommendations what each part is.
The practical move for distributors is the same as for retail. Keep the structure clean, then let search and adaptive filters do the heavy lifting. See AI search for B2B ecommerce.
What will not change about category pages
Some things are durable. Plan around them.
Search engines still need stable URLs.
A category page that ranks for a head term is an asset. It needs a permanent URL, crawlable content and consistent structure. Generated collections can supplement it, not replace it.
Undecided shoppers still browse.
A guided starting point beats a blank search box for many people, especially on mobile.
Merchandisers still need control.
Even with AI ordering, teams need to pin, boost and bury products for launches, promotions and stock positions.
In short, category pages survive as the stable backbone. The experience on top of them becomes dynamic.
Should you remove your category pages? A clear no
We want to be direct here, because the risk is real.
Do not delete ranking category pages.
You would lose years of search equity, and the traffic rarely returns to a generated replacement.
Do not hide navigation entirely.
Shoppers who do not know what to type need somewhere to start.
Do improve what happens on them.
Personal ordering, adaptive filters and conversational refinement can all sit on top of existing category pages without changing their URLs.
That approach gets the upside of AI search while protecting the SEO and browsing value you already own.
The headless angle, in plain terms
“Headless” simply means the storefront is separated from the commerce platform behind it. The platform holds products, prices and orders. The storefront decides what each page looks like.
Why does that matter here? Because a headless storefront can build a page from a search query as easily as from a category ID. As a result, generated collections become practical to launch and test.
However, headless is not required. Most of the changes in this forecast can sit on top of Shopify, BigCommerce, Magento or WooCommerce through a search API. See ecommerce search API for how that works.
How category pages affect SEO and AI visibility
Category pages also matter outside your site. AI answers and assistants increasingly summarize categories: “the best lightweight tents under $300,” for example.
A well-structured category page, with clear copy, accurate product data and consistent schema, gives those systems something reliable to read. For more, see AI visibility for ecommerce and product schema markup.
Meanwhile, the experience after the click still decides revenue. That is the core idea of SXO.
What usability research tells us
Research on category navigation is extensive. Baymard Institute’s homepage and category usability research is one of the most detailed public bodies of work on how shoppers move through navigation and product lists.
Nielsen Norman Group’s long-running study of ecommerce search found that faceted filtering moved from a nice-to-have to an expected feature, and that autosuggest, while widespread, was selected in only 23 percent of the cases where it was offered (NN/g).
The lesson for the future is consistent. Shoppers use several paths at once: search, filters, suggestions and browsing. The winning experience supports all of them rather than betting on one.
How to prepare in the next two quarters
Here is a practical sequence that protects what you have.
1. Inventory your category pages by value.
Note organic traffic, revenue and ranking terms for each. These are assets to protect.
2. Find your top unanswered intents.
Pull site search terms that do not map to any category. These are candidates for generated collections.
3. Clean the taxonomy.
Fix duplicate, empty and mislabeled categories. AI depends on this structure.
4. Add adaptive filters to high-traffic categories.
Start with the categories where shoppers filter most.
5. Test personal ordering on one category.
Use a control group, then expand if it wins. See A/B testing for ecommerce search.
Where the revenue comes from
It is worth being clear about why any of this pays. The gains come from three places.
Fewer dead ends.
When a category page answers the intent, fewer shoppers bounce to the search box or leave. That keeps more sessions alive.
Faster decisions.
Personal ordering and adaptive filters put a likely fit in the first row. Shoppers who find a match sooner tend to buy sooner.
New landing pages from real demand.
Generated collections turn popular unanswered searches into pages that can convert. Each one captures demand that previously had nowhere to go.
To turn those gains into a forecast, use the model in AI search revenue.
Metrics to watch
The bottom line on category pages
Category pages are not going away. Instead, they are being promoted from a fixed menu to a flexible starting point.
The stores that win will keep their stable, ranking category pages, clean the taxonomy behind them, and layer AI ordering, adaptive filters and conversation on top. The stores that lose will either ignore the change or tear down what already works.
For the wider picture, see our future of ecommerce forecast and AI shopping decisions.
Frequently asked questions
Will AI search replace category pages?
Not entirely. AI changes how category pages work by adding personal ordering, adaptive filters and conversational refinement, but stable category pages remain important for browsing, merchandising and search engine rankings.
Should I remove category pages to go AI-first?
No. Ranking category pages carry years of search equity, and deleting them usually loses traffic that generated replacements do not recover. Improve the experience on top of them instead.
What are dynamic or generated collections?
They are product sets assembled automatically from a shopper intent or popular query, such as “gifts for new hikers,” rather than built manually as a fixed category.
How does AI change filters on category pages?
Adaptive filters show the facets that matter for the current product set and query, such as width for boots or capacity for tents, instead of the same static list everywhere.
Does taxonomy still matter if shoppers search instead of browse?
Yes, arguably more. Retrieval, recommendations and AI assistants rely on clean category relationships to understand what each product is, even when shoppers never see the tree.
Can I have ecommerce search without category pages?
Technically, a headless storefront can assemble pages entirely from queries. In practice, most retailers should keep stable category pages for SEO and browsing and use dynamic collections as a supplement.
How do category pages affect AI visibility?
Well-structured category pages with accurate product data and consistent schema give AI answers and assistants reliable information to summarize when shoppers ask category-level questions.
What should I change first?
Clean the taxonomy, then add adaptive filters and test personalized ordering on your highest-traffic categories, measuring each change against a control.
How do I measure whether category pages are working?
Track organic sessions, category exit rate, filter usage, the rate of shoppers who search after browsing, and revenue per category session.
See what your category pages could become.
bCloud AI adds personal ordering, adaptive filters and intent-driven collections on top of your existing category pages, without touching the URLs that rank. See it on your catalog.
bcloud.ai
