Ecommerce SEO: The Complete Guide for Product Catalogs (2026)

1. 01

   Ecommerce SEO is catalog SEO. You rank through hundreds or thousands of templated category and product pages, so one fixed template or data field moves more traffic than any blog post you'll publish this year.

2. 02

   Category pages win broad queries like "linen shirts for men". Product pages win specific ones like "oatmeal linen shirt relaxed fit". Most stores plan keywords for only one of the two.

3. 03

   Filters, variants and pagination can spawn millions of near-duplicate URLs. On a large store, deciding what gets crawled and indexed comes before anyone touches a title tag.

4. 04

   The product data that ranks a page in Google (titles, attributes, identifiers, structured data) is the same data Merchant Center and AI shopping agents read. Fix it once and every channel gets better.

5. 05

   Each ranking earns fewer clicks as AI answers spread, while AI referrals convert better. We run SEO and AI search as one program, built on product pages that are complete and accurate.

Build the category tree from demand

Start from how people search, then fit merchandising around it. If shoppers type "women's running shoes", "trail running shoes" and "wide running shoes", you need three category or subcategory pages. Internal labels like "SS26 Performance Footwear" belong in the PIM and nowhere near your URLs.

A practical tree for a footwear store:

Google advises using descriptive words in URL paths and giving each distinct page its own path instead of a # fragment, since fragments aren't used for indexing (Google Search Central). Keep URLs short and lowercase, then leave them alone. Changing them later means redirects and months of waiting.

On Shopify, check how your theme links products from collections. Many themes output URLs like /collections/running-shoes/products/ridgeline-trail-gtx-black while the canonical points to /products/. That's thousands of internal links to an address you've asked Google not to index, and a small theme edit fixes it.

Internal linking for ecommerce

Google infers importance from internal links. Google also says "it's strongly recommended to link to all products that you wish indexed", using standard <a href> links rather than JavaScript click handlers (Google Search Central). Sitemaps and feeds are only a fallback.

Every product you want to rank needs at least one plain HTML link from a category page; a search box or a "load more" button won't get a crawler there.

For category-level detail, see our category page optimization guide.

What to optimize on every PDP

Baymard rates 52% of desktop and 62% of mobile product pages as mediocre or worse. A typical PDP imported straight from a supplier file shows why. Here it is before and after an Optimizer pass:

Images need several angles plus in-use shots, with alt text that describes the product. Reviews should be visible in the HTML, with aggregateRating in structured data. Write the product FAQ for shoppers first; whether marking it up with FAQ schema on product pages is still worth it is a smaller question.

Titles and headings

Put the product type, brand and defining attribute first. Google recommends descriptive, concise title elements and warns against keyword stuffing and boilerplate repeated across pages (Google Search Central). A "Buy X Online, Free Shipping" pattern on every PDP is that boilerplate, and Google tends to rewrite those title links.

Shopping works the same way. Merchant Center allows 150 characters, but shoppers usually notice only the first 70 or fewer (Google Merchant Center Help).

Original descriptions beat supplier copy

If 40 retailers publish the same supplier description, Google has little reason to pick yours. There's no penalty in the sense people fear; Google just shows one version, rarely the smaller store's. Rewrite descriptions to answer what a shopper asks before buying: who it's for, how it fits, what it's made of. Our guide on how to write product descriptions has templates by product type.

Complete attributes are keywords

Each attribute value is a query someone types. A product with no material value can't rank for "merino wool sweater", and an AI agent can't recommend it for "a sweater that doesn't itch". Google reports that retailers who added correct GTINs saw clicks rise 20% on average (Google Merchant Center Help).

A trap we see constantly is every variant sharing one GTIN because the supplier sent one barcode. Each variant needs its own.

Out-of-stock and discontinued products

• Temporarily out of stock: keep the page live and indexable, update availability in structured data and the feed, and show alternatives.
• Discontinued with a successor: 301 redirect to the new model.
• Discontinued with no replacement: redirect to the closest category, or return 404 or 410 if nothing is relevant. Redirecting hundreds of dead products to the home page gets them treated as soft 404s anyway.

Every empty attribute on a product page is a search query that page can never win.

The full PDP method is in Product Page Optimization.

Duplicates and canonicalization

Stores create duplicates through variants and tracking parameters, through products listed under several category paths, and through near-identical regional pages. Google ranks its signals: redirects and rel="canonical" are strong, sitemap inclusion is weak, and all of them should point to the same preferred URL (Google Search Central). When they disagree, Google picks for you. You'll see it in Search Console as "Duplicate, Google chose different canonical than user".

For variants, Google supports two models. Each variant can have its own URL (/t-shirt/green or /t-shirt?color=green), and with an optional parameter the URL without it should be canonical (Google Search Central). Shopify's ?variant= URLs follow that second pattern by default. Mark up variant families with ProductGroup, hasVariant, variesBy and productGroupID (Google Search Central).

Structured data for products

Merchant listing markup requires name, image and offers with a price and currency. Google also asks for gtin, brand, color, material, size, shippingDetails, hasMerchantReturnPolicy and aggregateRating where they apply (Google Search Central).

Google recommends using structured data and a Merchant Center feed together. Markup helps it read your pages, while the feed gives you direct control over which products appear and how often they update (Google Search Central). If your JSON-LD price comes from a cached template and the feed from the ERP, expect mismatch warnings in Merchant Center. Full examples are in Structured Data for Ecommerce, and feed setup is in Google Merchant Center.

International ecommerce SEO

For stores selling in the US, UK, Ireland, the Netherlands and Italy, each locale needs its own URL and hreflang annotations. Google's rules (Google Search Central):

• Declare alternates in HTML, HTTP headers or the XML sitemap. One method is enough.
• Every version lists itself and all the others. If two pages don't point to each other, the tags are ignored.
• Use ISO 639-1 language codes with optional ISO 3166-1 region codes, such as en-GB, en-IE, nl-NL, it-IT. A region code alone is invalid, and UK isn't a valid code.
• Add x-default for a country selector or fallback page.

Translation alone rarely ranks. Shoppers in Milan search "camicia di lino uomo", which a word-for-word translation often misses. Research keywords per market and localize attribute values and units, sizes included. A translated size chart still showing UK shoe sizes on the Italian store costs you in returns. See Ecommerce Localization.

Which ecommerce platform is best for SEO?

Every major platform can rank well. What differs is how much control you get over URLs, facets and languages, and what that control costs. We rarely advise replatforming for SEO alone; your data gaps move with you.

The best platform for SEO is the one whose data you can keep complete and whose templates you control. For platform-specific settings, see our WooCommerce SEO guide and our Magento SEO guide for Adobe Commerce.

• Category tree mirrors demand

  Every major query family has a matching category or subcategory URL

• Descriptive, stable URLs

  Lowercase words in paths, no fragments, no session IDs

• Every product linked

  Each indexable product has a plain HTML link from at least one category page

• Breadcrumbs on all pages

  Visible breadcrumbs plus breadcrumb structured data

• Keyword map by page type

  Head terms to categories, modifiers to subcategories or filters, specific queries to products

• Unique title tags

  No boilerplate duplicates; product type and brand first, then the key attribute

• Original product descriptions

  No supplier copy duplicated across the web

• Complete attributes

  color, material, size, gtin, brand and specifications filled for every product

• Product FAQ and reviews

  Real questions answered on the page; reviews visible and marked up

• Category intros

  50 to 150 words with links to subcategories

• Facet indexing rules

  Demand-led filters indexable; low-value combinations blocked; empty results return 404

• Pagination crawlable

  Unique URL per page, self-canonical, linked with <a href>

• Canonicals consistent

  Redirects, canonicals, sitemaps all point to the same preferred URL

• Variant handling

  One model chosen; ProductGroup markup in place

• Product structured data

  name, image, offers, plus gtin, brand, shipping and returns

• Merchant Center in sync

  Feed prices and availability match the landing pages

• Hreflang for each market

  Reciprocal tags, valid codes, x-default

• Localized keywords

  Research per market instead of translating English terms

• Stock lifecycle rules

  A rule for each case: out of stock, discontinued, replaced

• AI readiness

  Conversational attributes filled; use cases and intents covered; AI referrals tracked

Our ecommerce SEO audit expands this to 40 points, and the free AI Readiness Audit scores your product pages automatically.