Ecommerce SEO checklist: 40 checks for stores that want to rank in 2026

By Matija Konjić

Ecommerce SEO has more moving parts than any other kind: thousands of URLs, filters that multiply them, products that come and go, and a category structure that decides where the revenue queries land. The stores that rank are rarely the ones with the cleverest content; they are the ones that got the structure right and then built authority to the pages that sell.

This checklist is the one we run before any link building campaign for an online store, because links pointed at a broken structure are wasted. It has forty checks across seven sections, in the order that moves revenue, each with enough explanation to act on without a consultant. Work through it top to bottom; the early sections are where most stores lose, and the late sections are where they win.

How to use this checklist

Run it as an audit first, marking each check as pass, fail or unknown, then fix in order. Structure problems invalidate everything below them, category pages carry most of the revenue, product pages carry the long tail, technical issues leak authority, and content and links add the authority that decides rankings once the rest is sound.

A store with a clean structure, strong category pages and a modest link profile will outrank a store with a bigger catalog, more content and a tangle of duplicate URLs, which is the pattern we see in almost every audit.

Site structure and navigation

Structure decides what Google can reach and how authority flows from the homepage to the pages that convert.

• Every category is reachable within three clicks of the homepage. Deeper pages get crawled less and rank less.
• The main navigation links to the categories that earn revenue, and the footer carries the rest, so the strongest page on the site passes authority where it matters.
• Subcategories exist for every distinct buyer intent (“men’s running shoes” and “trail running shoes” are different queries) and each has one URL.
• Breadcrumbs are present on every category and product page, marked up with structured data, linking upward through the hierarchy.
• URLs are short, readable and stable, with no session IDs, no dates and no duplicate paths to the same category.
• Out-of-stock and discontinued products follow a rule: keep and mark unavailable if they will return, redirect to the closest category if they will not, never leave a 404 with links pointing at it.

Fix these before anything else, because every later section assumes Google can reach the pages and understand how they relate.

Category pages

Category pages are where “buy running shoes”, “leather sofas” and every other revenue query lands, and most stores give them a heading and a product grid.

• Each category has a unique title and H1 that names the product type in the words buyers use, with the modifier that distinguishes it from sibling categories.
• Each category carries useful introductory copy: a short paragraph above the grid that helps a buyer choose, and a fuller buying guide below it, written for the reader rather than for a word count.
• Sorting and filtering do not create indexable duplicates: filter URLs are canonicalized to the category or blocked, except for the handful of filter combinations with real search demand, which become their own landing pages.
• Pagination is handled consistently, with each page self-canonical and the products on page two reachable by crawlers.
• Category pages link down to subcategories and across to related categories in the body, with descriptive anchors.
• Category pages receive internal links from guides, blog posts and the homepage, since they rarely attract external links by themselves.
• Category pages are the primary target of link building, because that is where rankings turn into revenue.

A category page that passes all seven is a landing page in its own right, and it will outrank a competitor’s grid with twice the products behind it.

Product pages

Product pages carry the long tail and the purchase, and their problems are duplication and thinness.

• Each product has a unique title, H1 and description; manufacturer copy duplicated across hundreds of retailers ranks for none of them.
• Variants share one URL with a canonical unless a variant has its own search demand (a color or size people search for), in which case it gets its own page with its own content.
• Product structured data is complete: name, image, price, availability, reviews, brand and identifiers, validated without errors.
• Reviews are on the page and crawlable, since they add unique text and feed rich results.
• Images are compressed, named descriptively and given alt text that describes the product rather than the keyword.
• Each product links back to its category and to related products, so authority flows both ways and buyers can keep shopping.
• Thin or near-identical products are consolidated or noindexed, because a thousand two-sentence pages dilute the whole site.

Product pages do not need essays. They need to be distinct, complete and connected, which is a configuration problem more often than a writing one.

Technical SEO

The technical layer of ecommerce exists to stop the catalog multiplying into millions of worthless URLs and to make the ones that matter fast.

• Faceted navigation is controlled: crawlable facets are limited to those with demand, the rest are blocked in robots.txt or via nofollow and canonical tags, and the crawl stats in Search Console show Googlebot spending its budget on real pages.
• One canonical version of every page, with http, www, trailing slash and parameter variants all redirecting or canonicalizing to it.
• Core Web Vitals pass on mobile for category and product templates, because most ecommerce traffic is mobile and the templates repeat the same performance across thousands of URLs.
• The XML sitemap lists only indexable, canonical, in-stock URLs and updates when products change.
• Internal search result pages are noindexed, and so are cart, checkout, account and wishlist pages.
• Redirects are one hop, with no chains from old product URLs through old categories to new ones.
• Structured data covers products, breadcrumbs, the organization and, where relevant, FAQs on category guides, all validated.
• JavaScript-rendered content is visible to crawlers: product grids, prices and descriptions that load client-side need server-side rendering or pre-rendering.
• International versions use hreflang correctly and do not duplicate the same-language catalog across country folders without a reason.

Most of these are one-time fixes with lasting effect, and the first one, facets, is the single most common reason a large store underperforms its catalog.

Content that supports the catalog

Content on an ecommerce site has one job: to attract links and visitors that the category pages cannot, and to pass them on.

• Buying guides exist for every major category, answering the questions a buyer asks before choosing, and linking to the categories and products they discuss.
• Comparison and “best” content is honest and includes competitors’ products where that is what a buyer would expect, because that is what earns links and trust.
• Blog content targets queries with demand and links back to the catalog; a post that cannot link naturally to a category or product is a post for a different site.
• Seasonal and gift content is published early and refreshed yearly at the same URL, so it accumulates authority instead of starting over.
• Original data from your own sales or customers is published once a year, because it is the one asset journalists in your category will cite.

The ecommerce link building playbook covers how each of these content types turns into links; the checklist point is that content without a path back to the catalog is a cost rather than an investment.

Links and authority

Once the structure and pages are sound, links decide the rankings, and category pages almost never earn them without help.

• Referring domains are counted per category page, not per site, and the categories with revenue potential and the biggest gap to the current top five are prioritized.
• External links are built to guides, comparisons, data and tools that can earn them, and routed to categories through internal links.
• Category pages receive direct placements where the context fits: a buying guide on a relevant publication that links to the category it recommends is both natural and effective.
• Supplier, brand and partner links are claimed; manufacturers often maintain “where to buy” pages that will link to a stockist on request.
• Linking sites are vetted for traffic and relevance, since a link from a page nobody reads passes little, whatever its authority score.
• The twenty most-linked pages on the site link to the categories they support, which is usually the fastest win available.

That result came from a store with two decades of history and a competitor with twice its traffic; the work was guest posts and contextual placements aimed at category pages, with the catalog already in order. The full ecommerce SEO case study walks through the eight months. Our link building service runs the same program for stores that pass the first five sections of this list.

Measurement

Measure at the page type level, because the store-wide numbers hide where the problems are.

• Organic revenue by landing page type: categories, products, content, each month, from analytics.
• Rankings for the top fifty category queries, tracked weekly, with the category URL that ranks confirmed as the intended one.
• Indexed pages against intended pages in Search Console, so that the catalog is neither under-indexed nor bloated with facets.
• Referring domains per priority category, verified live and followed.
• Crawl stats, to confirm Googlebot spends its time on real pages.

A store that reviews those five numbers monthly catches structural regressions within weeks, which matters because platform updates, theme changes and new filters reintroduce the same problems this checklist removes. Run the list again after every major change to the store, and the rankings it earned stay earned.