How to Do E-commerce SEO: An SEO Guide for Online Stores

How is e-commerce SEO different from SEO for a corporate website?

E-commerce SEO differs from corporate website SEO in scale, templates and volatility: hundreds or thousands of products are generated from the same templates, filters and variants spin up new URLs on their own, and prices and stock change constantly. That’s why SEO for e-commerce is less about improving individual pages and more about managing templates, URL rules and product data.

The fundamentals still apply. HTTPS is essential for payment security and customer trust; Google described HTTPS as a lightweight ranking signal back in 2014, so it won’t carry your rankings on its own, but for a store its absence is a serious trust problem. Technical faults such as broken links, slow-loading pages and filters that don’t work should also be fixed before any content work starts. Measurement needs to be in place from the start too: until you can see which add-to-carts and purchases come from organic search, you can’t decide which categories to invest in. For the store’s page structure and pre-launch checks, see our article on SEO in web design.

How do you map keywords to category and product pages?

Keywords are mapped by search intent: generic product searches go to category pages, searches that include attributes go to subcategories or selected filter pages, brand and model searches go to product pages, and informational searches such as “how to choose” go to guides. Each query group is assigned to a single page; multiple pages targeting the same keyword end up competing with each other.

In practice, the research runs in this order: define the site’s goals and main themes, look at which keywords competing stores rank for, list the keywords by category and subcategory and prioritise them by search volume, add long-tail phrases and the seasonal demand visible in tools such as Google Trends, then monitor performance regularly and update the list. We explain the method in detail in our article on keyword research, and how to prepare for seasonal demand in advance in our article on seasonal SEO.

How do you optimise category pages?

Category pages are optimised with a unique title and H1, a short introduction that makes choosing easier, crawlable product and subcategory links, breadcrumbs and correctly implemented pagination. For generic product searches, the page that ranks is usually the category page, and it’s also the page that distributes internal links to the products.

• Title and H1: Should carry the name people actually search for the category by; avoid generic titles like “Products” or just the brand name.
• Category text: A short, genuinely helpful introduction above the product list that explains how to choose; blocks of keyword-stuffed text piled up at the bottom of the page don’t work.
• Product and subcategory links: Google recommends linking from menus to categories, from categories to subcategories and from there to every product; links should be <a> elements with an href.
• Breadcrumbs: Show the product’s category path and provide internal links to the parent categories.
• Elements that prompt action: Price, stock and shipping information should be visible in the list view too; product titles and category navigation should clearly lead the visitor to the next step.

For pagination, Google recommends giving each page its own URL such as ?page=2, linking the pages to each other sequentially, and not setting the first page as the canonical for the whole series; it no longer uses rel=next and rel=prev. If you use a “Load more” button or an endless product list, the same products must also be reachable through crawlable pagination links, because Google’s crawlers don’t press buttons and generally don’t trigger JavaScript functions that wait for a user action.

How do you use original content, images and reviews on product pages?

A product page gets stronger with an original description that answers the buyer’s decision questions instead of repeating the copy the manufacturer gives everyone, optimised images that show the product from different angles, and genuine customer reviews. Among dozens of stores using the same manufacturer copy, Google needs a reason to put your page ahead.

Google says some duplicate content is normal and doesn’t violate its spam policies; but if the same description appears in many stores, the rest of the page makes the difference. Content copied from other sites without adding any original value, on the other hand, is defined in Google’s spam policies as scraped content. In large catalogues, starting the original descriptions with your best-selling and most-searched products is a practical way to prioritise.

• Title pattern: Brand, model and distinguishing attribute (colour, capacity, size) in that order; the same template for every product.
• Description: Use case, dimensions, material, compatibility, care and what’s in the box; the questions customer service hears most often should be answered on the product page.
• Images: Photos taken from different angles and in real-use settings; descriptive file names and alt text, modern formats and defined dimensions. The main product image is often the largest element on the page and shouldn’t be lazy-loaded.
• Reviews: Reviews from verified buyers add original content and trust to the page; rating markup should reflect only genuine reviews visible on the page. Fabricated reviews create both legal and reputational risk.
• Consistent information: Price, stock, shipping cost and return terms should be the same on the page, in the structured data and in the product feed.

We cover the technical side of images in our article on image SEO, and how a product page makes buying easier in our article on e-commerce conversion rates.

How do you manage filters, variants and parameterised URLs?

Filters, variants and parameterised URLs are managed by opening the few combinations with real search demand to indexing and keeping the rest out of crawling. Filters are indispensable for visitors; the problem isn’t the filter itself but the fact that once colour, size, price range and sorting options combine, a single category can generate thousands of URLs, most of which show the same products in a different order.

Google’s documentation on faceted navigation says these URLs cause two problems: the crawler spends time crawling useless URLs, and the discovery of new, valuable pages slows down. For filter URLs that don’t need to be indexed, Google recommends blocking them with robots.txt or implementing the filter with a URL fragment (#); it generally finds rel=canonical and nofollow less effective in the long term. For filters that will be indexed, it recommends the standard & separator, always listing filters in the same order, and a 404 status code for combinations that return no results.

For variants, Google supports both approaches. If all the options are on one page and selected with a parameter, the ProductGroup markup points to the base URL without a variant selector; if the variants are on separate pages, each page carries its own markup. Either way, every variant should have a unique SKU or GTIN and be directly selectable through its own URL; the hasVariant, variesBy and productGroupID properties define this relationship. For a model people search for in a specific colour, giving that variant its own page makes sense; for size options nobody searches for separately, a single page is enough.

What should you do with out-of-stock and discontinued products?

The page of a temporarily out-of-stock product should stay live, with its stock status shown correctly both on the page and in the structured data; a permanently removed product should be redirected with a 301 to a direct equivalent if there is one, and should otherwise return a 404 or 410. Deleting the page every time stock runs out, or redirecting every old product to the homepage, throws away the visibility you’ve earned.

According to Google’s documentation, 404 and 410 are treated the same way, and a previously indexed URL is dropped from the index. Its site move guide also says that redirecting large numbers of old URLs to a single irrelevant destination such as the homepage can confuse users and may be treated as a soft 404. “Product not found” pages that return 200 despite having no content fall into the same category and use up crawl resources.

Why do Product structured data and Merchant Center matter?

Product structured data and Merchant Center pass a product’s name, price, stock status and shipping and return terms to Google in a machine-readable form; this data is what allows a product to be shown in search results enriched with price and stock information, and on Google’s shopping surfaces.

Google recommends using both together: structured data on the page helps Google understand the page and show it as a rich result, while a Merchant Center feed gives you more control over the timing of updates and lets you share data that isn’t on the site. According to Google, Merchant Center is required to appear in the Shopping tab but not to appear in general search results.

• Merchant listings: Only pages where the product can be bought directly are eligible. name, image and offers are required; the Offer must contain a price greater than zero and a priceCurrency in ISO 4217 format.
• Product snippets: For product pages that don’t sell directly, such as review and comparison pages.
• Stock status: Values such as InStock, OutOfStock, BackOrder, PreOrder and Discontinued must match the status shown on the page.
• Shipping and returns: Google recommends providing these policies business-wide with Organization markup, and at Offer level (shippingDetails, hasMerchantReturnPolicy) only for products with different terms.
• Monitoring: The Merchant listings and Product snippets reports in the Shopping section of Search Console show markup errors separately.

Structured data doesn’t guarantee a rich result; Google expects the markup to match the information visible on the page. Checking the required fields of the Product markup on a product page with our Product schema checker is a practical way to confirm that match. A product that shows a discounted price on the page but carries the old price in its schema can lose its enhanced display because of that mismatch. We cover the implementation details in our structured data guide.

How do you manage crawl budget and page speed for large catalogues?

In large catalogues, crawl budget is managed by not making Google spend its time on worthless URLs: duplicate and unnecessary URLs are cut down, removed products are closed off with the right status code, the sitemap is kept up to date and pages respond quickly. Google’s crawl budget guide is written specifically for sites with more than a million unique pages whose content changes once a week, sites with more than ten thousand pages whose content changes daily, and sites where a significant share of URLs sit in Search Console as “Discovered - currently not indexed”.

For most small and medium-sized stores, crawl budget isn’t a real constraint; but uncontrolled filters can turn even a small store into a huge site in a crawler’s eyes.

1. Manage your URL inventory. Block the unnecessary URLs generated by filter, sorting and session parameters from crawling with robots.txt; Google recommends robots.txt rather than noindex for this, because a page carrying noindex still gets crawled.
2. Consolidate duplicate content. Bring the different paths to the same product together under a single canonical URL.
3. Give removed products the right code. Return 404 or 410 for permanently removed pages and weed out soft 404 pages.
4. Keep your sitemap up to date. List only the canonical product and category URLs you want indexed, with an accurate lastmod value.
5. Shorten redirect chains. Long chains hurt crawling; old URLs should reach their final destination in a single hop.
6. Make pages load fast. Pages that respond quickly and can be cached put less load on both the crawler and the server.

We measure speed at template level: LCP, INP and CLS are reviewed separately for each of the category, product, cart and checkout templates. Because Google uses the mobile version of a site for indexing, the product description, reviews and structured data must also be complete in the mobile template. On a product page, the LCP element is usually the main product image; INP suffers from interactions such as filter and variant selection, and CLS from campaign banners that load late. Compressing images and converting them to modern formats, a CDN, browser caching, minifying JavaScript and CSS files, and solid hosting are the standard tools; the metrics are covered in detail in our Core Web Vitals guide. For large catalogues that hit the limits of an off-the-shelf platform, custom e-commerce software or e-commerce web design are also worth considering.

How are AI summaries and marketplaces changing competition in shopping searches?

In shopping searches, AI summaries have started to bring the comparison a user would make across several tabs together in a single answer, which makes it easier for stores that offer complete, consistent product data, content that spells out product features clearly and trustworthy reviews to make it into that answer. Marketplaces remain strong in generic product searches, while an independent store’s advantage lies in expertise, brand and niche searches.

Google began rolling out AI Mode and AI Overviews in Türkiye on February 18, 2026. The AI Mode shopping experience Google announced for the US in May 2025 combines Gemini with the Shopping Graph, which Google says contains more than 50 billion product listings, and works out product criteria by splitting a query into several searches at once (query fan-out). The scope of these shopping features can vary from country to country, but the direction is clear: the more complete and consistent your product data, the easier it is for AI to carry it into an answer.

• Keep product data complete: Brand, model, GTIN, dimensions, material, colour, stock, shipping and return information in your Merchant Center feed and structured data.
• Spell out features in the text: The answers to questions like “Is it waterproof, what foot type is it suited to, what’s in the box?” belong in the text, not inside an image.
• Create comparison and guide content: Guides that answer “which one should I choose?” questions are the realistic way to compete with marketplaces in generic searches.
• Don’t block AI search bots: OpenAI, for example, recommends allowing OAI-SearchBot in robots.txt so that sites can appear in ChatGPT’s search features; CDN and firewall settings shouldn’t block these bots either.

Marketplaces can complement a store as a sales channel, but a customer won on a marketplace often remains the platform’s customer. Growing brand searches, returning customers and a list of customers who have agreed to hear from you on your own site builds a long-term asset. We cover how AI summaries work in general in our AI Overviews and AI Mode article.

We run product page optimisation, category SEO and structured data work in the same monthly plan in our e-commerce SEO packages; a free analysis is a good starting point for seeing where your store stands today.