What ecommerce SEO actually is
Ecommerce SEO is the work of making your store discoverable so more customers find your products through Google and other search engines. It differs from SEO for a content site in concrete ways: your most important landing pages are products and categories rather than articles, search intent is mostly commercial, and the platform under the hood drives more of the result than on a regular site. That is why SEO work for a Shopify merchant looks different from SEO work for a B2B wholesaler on Adobe Commerce with a gated catalogue.
This guide covers the basics and then goes into what actually moves traffic: product pages, category pages, technical SEO, structure, B2B-specific challenges, and how classical SEO connects to the new AI search landscape. The tone is advisor, the position is platform-neutral. We work across four platforms and have no interest in pretending one of them is the right answer for everyone.
The four pillars of ecommerce SEO
SEO work is usually split into four areas. For ecommerce the weight between them is somewhat different from a content site.
Technical and platform. How fast the site loads, how product pages render, how URL structure and hreflang are set up, how the site handles faceted navigation without creating duplicate content. For ecommerce, technical SEO is usually the heaviest pillar because the platform shapes how products and categories are exposed to Google.
Structure and internal linking. How categories, subcategories and products connect, how deep a product sits from the home page, how navigation and breadcrumbs support both search engines and users. A rule of thumb is that any product on a mid-sized store should be reachable in three to four clicks from the home page.
Content. Product titles, descriptions, category copy, guides and blog posts that cover informational searches earlier in the customer journey. The challenge for ecommerce is that product copy is often templated or supplier-provided, leaving large parts of the catalogue ranking weakly.
Authority and links. External links from credible sites and brand mentions. For ecommerce a lot of this comes naturally from PR, media reviews and supplier partnerships.
Keyword research for ecommerce
Keyword research on an ecommerce store differs from a content site in one important way: you have to cover both commercial and informational intent, and you need to map them to different page types. A commercial query like "white running shoes men's" belongs on a category or product page. An informational query like "how to choose running shoes" belongs on a guide or blog post.
In practice: start with your most important categories and products, list the exact words customers use, and separate out the informational keywords that can feed a content strategy. For most merchants the long-tail (three to five word phrases) is the biggest missed opportunity, since competition is lower and intent is more specific.
Technical SEO that actually moves the needle
Technical SEO is a broad topic. For ecommerce it is a small set of issues that keep coming up as real problems.
Server-rendering of product pages. If your product specs, prices and reviews only render client-side in the browser, Google and AI bots see a near-empty page. Open a product page with JavaScript disabled. If the price, specs and reviews disappear, you have an indexing problem that no amount of content work compensates for.
Core Web Vitals. Google measures loading (LCP), interactivity (INP since March 2024) and visual stability (CLS). For ecommerce LCP is usually the heaviest because hero and product images are large. Practical action: optimise images, use modern formats like WebP or AVIF, and put reasonable limits on third-party scripts.
Indexing control. Faceted navigation often creates thousands of URLs combining filters (size + colour + price). They must not be indexed separately. The solution is canonical tags, robots-meta noindex on filter combinations, and in some cases robots.txt blocks. The goal: Google indexes your category and product pages, not 50 000 filter combinations.
Structured data. Implement Product schema with Offer, AggregateRating, Review, GTIN or MPN, shipping and return data on every product page. BreadcrumbList at every level. Organization schema with sameAs links to Wikidata, LinkedIn and Trustpilot. This affects both classical search results and AI search.
Hreflang. If you run multiple markets (sv, no, dk, en), hreflang tags must match exactly across all locale versions. Wrong hreflang is one of the most common reasons Swedish merchants are invisible in other Nordic markets.
Product pages that rank
The product page is the most important landing page on an ecommerce store. It should rank on the product name, brand plus model, and long-tail combinations of attributes. To make it work:
Unique product description per product. Templated descriptions from the supplier are one of the biggest SEO problems on most ecommerce sites. Write your own copy where you can, at least on your top sellers.
Structured attributes, not only free text. Material, size, certifications, delivery time — these should live as structured data, not just in the body copy. That makes them machine-readable for Google and AI systems.
Reviews. A Search Engine Land analysis in early 2026 of 1 000 products that AI systems recommend found a median of 156 reviews per product. Below roughly 150 reviews, AI search visibility collapses. For classical SEO, reviews also drive star ratings in SERPs. Use Yotpo, Trustpilot, Judge.me or another solution that exposes reviews in HTML, not just in iframes.
Clear titles and meta descriptions. The title should contain product name, brand and key attributes. The meta description should sell the click, not repeat the title.
Image optimisation. File names, alt text, compression. Image search drives more ecommerce traffic than many merchants realise, especially in fashion, home and interior categories.
Category pages and information structure
Category pages are often the underrated SEO asset on an ecommerce store. They rank on broader keywords ("hiking shoes men's", "ergonomic office chairs") that often have higher volume than individual product names.
For category pages to rank they need to be treated as real content: an H1 that matches search intent, a short introduction explaining the assortment, correct breadcrumbs, internal linking to subcategories and related products. Templated or missing category copy is one of the most common reasons categories do not rank.
For faceted navigation: decide which filter combinations actually have search volume (for example "hiking shoes men size 43") and turn them into indexable, optimised pages. Other filter combinations should be noindexed or canonicalised to the parent category.
SEO on Shopify
Shopify has a strong starting point: the default Liquid storefront is server-rendered, the site loads fast, and Product and BreadcrumbList schema come built into most themes. Shopify Catalog also syndicates products to ChatGPT, Google AI Mode, Microsoft Copilot and Perplexity by default.
Harder parts: the URL structure is partially locked ("/products/", "/collections/"), which makes some sites feel formulaic. Large catalogues need careful collection-page handling to avoid duplicate content. App work can drag down LCP if you are not selective. For more specific guidance, see our Shopify in Sweden guide and Shopify B2B guide.
SEO on Shopware
Shopware gives more control over URL structure, data models and frontend than Shopify. Server-rendering is the default. The platform has strong product attribute handling, which makes it easier to expose structured data and faceted navigation without losing control.
Harder parts: Shopware demands more from the development team to get the basics in place, and custom storefronts (especially headless) can lose server-rendering if the implementation is not deliberate. Shopware's rule engine for prices and campaigns is powerful but requires careful indexing control when prices are customer-specific.
SEO on Adobe Commerce with Hyvä
Adobe Commerce with the Hyvä frontend ships fast, server-rendered pages and strong Core Web Vitals scores. URL structure, hreflang, robots.txt and canonical handling are all flexible — which is both the strength and the trap. The platform gives you access to everything, but you have to configure it.
Harder parts: faceted navigation and layered navigation can generate huge amounts of URLs that need active indexing control. The default Luma frontend is heavy and hurts Core Web Vitals — switch to Hyvä if performance is a problem. See our Magento performance guide for a deeper walkthrough.
SEO on Norce headless
Norce is API-first and most often used headless, which makes SEO a frontend question rather than a platform question. With Frntkey as the composed frontend you get full control over server-rendering, performance and URL structure.
Harder parts: server-rendering is not free in headless architecture. It has to be built into the frontend implementation. If server-rendering is not in place, product pages become invisible to Google and AI bots. This is the single most common pitfall we see on headless projects.
SEO for B2B ecommerce
B2B merchants face a number of SEO challenges that B2C usually avoids.
Gated catalogue. If your entire product range sits behind login, Google sees nothing. The solution is usually a public product catalogue with list prices and a separate logged-in view with customer-specific prices. That opens SEO without losing the B2B logic.
Customer-specific pricing. With customer-specific pricing, no price shows up in SERPs via Product schema. That is fine. Display list price in structured data and communicate customer-specific pricing as an option in the body text.
ERP-driven product data. Products, inventory and variants often come from the ERP. For SEO this means the data quality in the ERP directly affects SEO results. For a deeper walkthrough see our ERP integration guide.
Long-tail B2B keywords. B2B searches are often extremely specific ("hydraulic hose DN12 1m"). Long-tail is where conversion lives. Invest in structured attribute data and faceted navigation that covers those combinations.
SEO and AI search in 2026
Classical SEO and AI search are not two separate disciplines. Research from Ahrefs and Profound shows that around 99% of citations in Google AI Mode come from the top 20 organic results. That means classical SEO is the precondition for visibility in AI search. AI search optimisation is a sharpening on top, not a replacement.
What has changed since Google AI Overviews launched in Sweden on 9 May 2025: informational queries are losing clicks (Ahrefs measured a 58% drop in position-one CTR when an AI Overview appears), while transactional and product queries are largely intact. That is why ecommerce merchants should defend product and category pages and adapt informational content to passage-based retrieval at the same time.
For a deeper walkthrough of AEO, GEO and llms.txt, see our guide to SEO for AI search. It covers exactly what to do and what to skip when AI search engines enter the picture.
Common mistakes on Swedish ecommerce sites
The recurring problems we see on Swedish merchants:
Thin product copy or templated descriptions. Almost every store has this on the long tail of the catalogue. Prioritise top sellers and high-margin products first.
Headless builds without server-rendering. A modern React or Vue frontend without proper SSR makes Google and AI bots see empty pages. Fix this before investing in any other SEO work.
Wrong hreflang on multi-market setups. If hreflang tags point incorrectly between SE, NO and DK, all three markets get poor indexing. Audit when you expand.
Too many indexable filter combinations. Thousands of thin filter URLs cut authority on your real category pages. Decide which combinations actually have search volume, index those, block the rest.
Missing or wrong schema. Product schema without Offer or AggregateRating, or with the wrong currency, is very common. Audit with Google's Rich Results Test on your top sellers.
No or few reviews. Reviews drive both classical SEO via star ratings and AI search via product recommendations. Set up a system and ask for reviews actively.
How we work with ecommerce SEO
Nordic Web Team treats SEO as part of normal ecommerce delivery, not as a separate workstream. We audit how your store renders to Google and AI bots, fix technical foundations, prioritise product and category pages, implement structured data, and build link and content strategies that fit your business.
We work platform-neutrally across Shopify, Shopware, Adobe Commerce with Hyvä and Norce. The right starting point is usually a short SEO audit that maps your current state against the priority list in this guide. From there it becomes a normal delivery — platform work, content work and measurement that go hand in hand.
SEO drives traffic, but conversion drives revenue. Once SEO is sending visitors to product and category pages, our conversion rate optimisation guide covers what to test next on product pages, checkout, and category listings.