By sector

Large catalogues fail in ways small sites never do

Forty thousand products generated by a system that also generates a page for every combination of three filters. The problem is not the products. It is the several hundred thousand pages nobody intended to create.

Analyse your store Discuss an assessment

Ecommerce sites have a structural problem that content sites do not. They generate pages automatically, at scale, from a template, and the rules that generate them were written by somebody who was thinking about merchandising rather than about how many addresses would result.

Add faceted navigation — colour, size, price band, brand, availability — and a catalogue of forty thousand products becomes a site with several hundred thousand reachable addresses, most of which differ from each other trivially and none of which anybody decided to publish.

Search engines allocate finite attention to any site. Spending it on two hundred thousand filter combinations means spending less of it on the category pages that actually earn revenue.

Focus

What matters on a retail site

The framework is the same as for any website. Where it points is not, and these are the areas where retail sites consistently differ.

Category architecture

Whether categories match how buyers actually search
Depth of the hierarchy and how far products sit from the top
Categories competing with each other for one intent
Whether category pages contain anything beyond a product grid

Faceted navigation

How many addresses filter combinations generate
Which combinations are reachable, indexable or both
Whether valuable facets are deliberately available to search
Canonical handling across filtered variants

Product pages

Whether descriptions are original or supplied by the manufacturer
Product markup accuracy, particularly price and availability
Whether specifications are structured or embedded in prose
Review presentation and its markup where genuinely present

Catalogue lifecycle

Handling of out-of-stock and seasonal items
Treatment of discontinued products with accumulated value
Variant handling across size, colour and configuration
Whether retired products redirect sensibly or return errors

Commercial trust

Delivery and returns information at the point of decision
Whether total cost is clear before the basket
Contact route for pre-purchase questions
Trust signals present where the decision is made

Performance at scale

Image handling across a large catalogue
Category page performance with many products loaded
Mobile behaviour on the browse-to-basket path
Whether product content depends on JavaScript execution

A retailer's crawl budget is finite. Spending it on two hundred thousand filter combinations means spending less of it on the twelve category pages that earn the revenue.

Pattern

Where catalogue sites lose most

The single largest recurring problem is category pages that are nothing but a product grid. They rank for the most commercially valuable searches in retail — the category terms — and they contain no information that would help a search engine understand what the category is or a buyer decide what to choose within it.

The second is faceted navigation left to generate whatever it generates. This is not solved by blocking everything; some filter combinations correspond to real searches with real demand, and blocking them wholesale forfeits that. It is solved by deciding deliberately which combinations deserve to exist.

The third is product content copied from the manufacturer, which places you in competition with every other retailer carrying the same line, all of them using identical text.

Commonly surfaced

Category pages with no content beyond the grid
Filter combinations generating six-figure address counts
Product descriptions identical to competitors’
Structured data prices that disagree with the page
Discontinued products returning errors, not redirects
Variants competing with each other as separate pages
Delivery cost first revealed at the basket
Images served far larger than they are displayed

Example finding

A representative finding

This is an illustration of the report format rather than a result from a real assessment.

High

Filter Combinations Are Generating Indexable Pages At Scale

What we found: Category pages produce a distinct address for each combination of filters applied. Sampling across three categories indicates the pattern generates well into six figures of reachable addresses. These pages carry no canonical relationship to their parent category and are not excluded from indexing.
Why it matters: Search engines allocate finite attention to any site. Spending it on filtered variants of a category, most of which show a handful of products and differ from each other trivially, reduces the attention available for the category pages themselves — which are the pages that earn revenue. It also creates competition between a category and hundreds of near-copies of it.
General direction: Decide deliberately which filter combinations correspond to genuine search demand and should exist as pages, and exclude the remainder from indexing while keeping them usable for shoppers. This is a decision about which combinations have commercial value, not a blanket technical rule.
How IXSEO can help: Technical SEO, with SEO Strategy where category structure needs revisiting.

Frequently asked questions

We have forty thousand products. Can you assess that?

Not exhaustively in a passive snapshot, and exhaustive assessment is rarely what a large catalogue needs. Problems on catalogue sites are template problems: they affect every product of a type, not individual items. Assessing a representative sample across each template establishes the pattern, and paid engagements include a full crawl where the scale warrants it.

Should we write unique descriptions for every product?

For products where the manufacturer's description is shared with fifty other retailers and the purchase is considered, yes — that is where original content earns its cost. For a catalogue of technical components bought on specification, the return is close to zero and the effort is better spent on category pages. The assessment identifies which category you are in.

How do we handle out-of-stock and discontinued products?

It depends on whether the product is returning. Temporarily unavailable items should generally stay, clearly labelled, with alternatives offered. Permanently discontinued items with accumulated value should redirect to the closest equivalent rather than to the home page or a general category, which is the most common mistake.

Does structured data affect our Shopping listings?

Product markup supports free listings and rich results, and its accuracy matters more than its presence — price and availability that disagree with the page cause problems rather than solving them. Paid Shopping campaigns run from your feed, which is a related but separate data source with its own quality issues.

Our platform limits what we can change.

Most do, and the assessment accounts for it. Where a finding cannot be addressed within your platform, that is stated alongside what can be done instead. There is usually more room than the platform's documentation suggests, and occasionally the honest answer is that a constraint is genuinely binding.

ServiceTechnical SEOWhere faceted navigation and indexation problems get resolved.ServiceWebsite Migration SEOReplatforming a store without losing what the catalogue has earned.ModuleConversionThe browse-to-basket path, assessed for obstacles.

Start with the store you have

The snapshot examines the home page and a sample of category and product pages. On catalogue sites, the pattern it finds usually applies at scale.