E-commerce SEO: Where to Start Optimizing Your Catalog
An online store differs from a regular site in one respect: page count. A thousand product listings plus filters produce tens of thousands of URLs, and without upfront architecture the crawl budget burns on junk combinations while the categories that matter stay unindexed.
Category architecture: build for demand, not warehouse logic
Catalog structure should mirror how people search, not how the warehouse classifies inventory. If shoppers search "men's winter jackets" but the site only has "Outerwear → Jackets" with no split by gender or season, that page competes on the broad term instead of the exact one and loses to niche competitors with tighter categories.
Nesting depth matters for crawling: a category four clicks from the homepage gets crawled noticeably less often than one at the second level. The rule of thumb — three clicks from homepage to product page — still holds up as a working target.
Breadcrumbs marked up with BreadcrumbList aren't just navigation convenience — they're extra internal linking that passes weight from top-level categories down to nested ones, and a signal to Google about where the page sits in the site hierarchy.
Product pages: uniqueness through data, not manual rewriting
The core problem with 5,000-to-10,000-SKU catalogs is identical supplier descriptions copied verbatim across dozens of competing stores. Rewriting every one by hand isn't realistic, so uniqueness gets assembled from structural elements instead: a spec comparison table, a "bought together" block, real customer Q&A, and locally shot photos instead of stock images.
Schema.org Product markup with price, availability, and aggregateRating fields is required to earn a rich snippet with price and rating directly in results — that lifts CTR even with no ranking change. Google Merchant Center separately validates this markup for Google Shopping placement.
An out-of-stock product shouldn't just disappear from the site — a 404 on a page that has already accumulated link equity and click history throws that asset away for good. The right practice: keep the page live with an "out of stock" label, surface similar products, and offer a back-in-stock notification signup.
Filters and duplicates: the catalog's biggest technical risk
Every filter combination — color, size, brand, price — generates a distinct URL with GET parameters. Left uncontrolled, Google indexes thousands of variations of what is functionally the same page, diluting link equity and creating duplicate titles and descriptions.
The working practice: a canonical URL (rel=canonical) on filtered pages points back to the base category page without filters. Combinations with genuine commercial demand — "men's winter blue jackets" with real search volume — get pulled out into their own indexable page with a unique title instead of being left as a stray parameter.
In robots.txt, it makes sense to block utility sort and pagination parameters (?sort=, ?view=) via Disallow, while leaving parameters that form commercially meaningful URLs open and actively optimized. Conflating the two is a common mistake that costs a store both crawl budget and traffic.
Pagination and infinite scroll
Paginated navigation on categories with hundreds of products needs to be crawlable: links to pages 2, 3, 4 as plain <a href> elements, not only a JavaScript "Load more" button that never changes the URL. If the bot can't see a link, it never sees the products past the first screen.
Infinite scroll with no pagination reflected in the URL is a common reason a thousand-product catalog ends up indexed at only 40-60 first-screen positions. The fix is adding a parallel paginated version with a page query parameter that both the pagination controls and the sitemap link to.
For an e-commerce sitemap.xml, split it into separate files by type — sitemap-categories.xml, sitemap-products.xml, sitemap-static.xml — under a sitemap-index.xml index. That makes Search Console diagnosis much faster: you immediately see which section has the indexing problem.
Speed and mobile: more critical here than for a landing page
On a category page with 40-60 listings and filters, every extra analytics script or unoptimized image preview hits LCP directly. Use WebP with lazy loading for products below the fold and eager loading for the first 4-6 visible listings.
Mobile traffic on Uzbekistan e-commerce sites typically runs above 70%, so Google's mobile-first indexing means: if filters are hidden behind a broken button on mobile or product cards get clipped, that's a primary ranking problem, not a secondary one.
First-month priorities
In order: technical foundations first (canonicals, robots.txt, sitemap-index, eliminating duplicate titles), then category structure built around real demand from Search Console and keyword data, then content for the top 200 products by margin or sales volume, and only after that expanding the long tail of smaller categories.
The metric for tracking early progress isn't rankings — it's the "Page indexing" report in Search Console: the share of submitted-sitemap URLs showing as "Indexed." Moving that share from 30% to 70% in the first month is a realistic, measurable target for a catalog nobody has optimized before.