Category pages are the ones that earn
The instinct on a new store is to optimise product pages, because products are what you sell. On most stores that is the wrong priority.
The reason is in how people search. Someone typing "waterproof walking boots" is not looking for one specific boot — they want the range. A category page showing thirty boots answers that query. A single product page does not.
Product-page searches are narrower and later: a model name, a part number, a size. Valuable, high-converting, and far fewer of them.
So the structural rule is: category pages compete for the broad, high-volume terms; product pages compete for specific ones. A store with excellent product pages and thin, templated category pages has optimised the small half of its opportunity.
Practically, that means a category page needs to be a page, not a grid. A heading that uses the term people search, a short piece of genuinely useful copy about the range, and internal links to related categories — not 400 words of keyword-stuffed filler pushed below the fold where nobody reads it.
Faceted navigation is the biggest technical risk
Filters are where ecommerce sites generate URLs faster than anyone can manage them.
Colour, size, brand, price, sort order — each filter multiplies. Five filters with a handful of options each can generate thousands of URLs, most showing near-identical content, all competing with the category page they came from.
The decision is per facet, not site-wide, and it comes down to one question: does anyone search for this combination?
- "waterproof walking boots" — a real query with real demand. That filtered view may deserve to be indexed.
- "walking boots sorted by price descending, page 3" — nobody searches this. It should not be indexed, and ideally should not be crawlable.
The usual handling is to let genuinely searched combinations be indexable with their own titles and copy, and to canonicalise or block the rest. What matters is that somebody decided, rather than the platform generating whatever it generates.
Left alone, facets produce a site where most crawlable URLs are worthless, and the pages that matter compete with thousands of variants of themselves.
Product variants
The same product in six colours is one of three things, and the choice has consequences:
One page with variant selectors. Usually correct. All authority consolidates, one URL competes, and the selector handles the rest.
Separate pages per variant, canonicalised to a main version. Reasonable when each variant needs its own URL for merchandising, but only one should compete in search.
Separate indexable pages per variant. Only justified when people genuinely search for the specific variant — "red Nike Air Max 90" as a distinct query with its own demand. Rare, but real in fashion and some parts.
The failure mode is the default: six near-identical pages, each with the same description and a different colour swatch, all competing.
Out-of-stock and discontinued products
The most common ecommerce SEO decision, made badly by default. Four options:
| Situation | Do this |
|---|---|
| Temporarily out of stock | Keep the page live. Show stock status, offer alternatives |
| Permanently gone, replacement exists | 301 to the replacement |
| Permanently gone, no replacement | 410, or keep the page live if it still earns traffic |
| Seasonal, returning | Keep it live. Do not delete and recreate annually |
Deleting a product page the day it goes out of stock throws away every ranking, link and piece of history it accumulated. If it returns next season, you start again from nothing.
A product page that still receives search traffic is worth keeping even when you cannot sell the item — it can route people to alternatives instead of returning a 404 to a warm visitor.
What to fix, in order
Diminishing returns set in fast, so the order matters more than the list.
- Indexation. Find out what is actually indexed versus what should be. On a store with facets this is usually where the surprise is.
- Category pages. The high-volume terms live here. Titles, headings, copy, internal links.
- Faceted navigation. Decide per facet. Stop the URL explosion before optimising anything downstream.
- Product page templates. Fix the template once and every product improves. Unique descriptions beat manufacturer copy, which is on every competitor's site too.
- Site speed. Ecommerce pages carry heavy images and many scripts; this affects conversion as directly as ranking.
- Structured data. Product schema with price and availability, kept accurate.
The order is deliberate. Optimising product copy while facets generate ten thousand duplicate URLs is decorating a room that is on fire.
Frequently asked questions
Should I optimise category or product pages first?
Category pages, on most stores. Broad terms like "waterproof walking boots" are answered by a range, not one item, and that is where the volume sits. Product pages compete for narrower, later-stage queries.
How do I handle out-of-stock products?
Keep the page live if the item is returning or still earns traffic, and show alternatives. Redirect to a genuine replacement if one exists. Only return 410 when the product is permanently gone with no replacement and no traffic.
Should filtered pages be indexed?
Only combinations people actually search for. "Waterproof walking boots" may deserve indexing; "sorted by price, page 3" does not. The decision is per facet — the danger is letting the platform generate thousands of crawlable variants by default.
Do I need unique descriptions for every product?
Manufacturer descriptions appear on every competitor's site, so they give you nothing to rank on. Writing unique copy for your best-selling and highest-traffic products first is usually a better use of effort than rewriting a 10,000-product catalogue.
Find out where your store's opportunity actually is
Ecommerce SEO gets expensive when effort goes to the wrong layer — rewriting product copy while faceted navigation quietly generates thousands of duplicate URLs competing with the categories that could rank.
If you want to know where the demand is and which pages could realistically capture it, the E-Commerce Opportunity Report measures the demand around your catalogue and where you stand against it — $197, one-off, no retainer.