Technical SEO Consultant, Sweden

Webshops, headless builds and catalogs fed from the ERP

Swedish ecommerce runs on a mix of hosted platforms such as Shopify, open-source stores such as WooCommerce, Nordic commerce platforms and headless builds where a JavaScript framework sits in front of a product API. Each stack fails in its own way. Hosted platforms impose URL patterns you cannot fully change; headless builds may ship an almost empty HTML shell that only fills after scripts run; plugin-heavy stores often output duplicate tags.

Behind many of these stores sits an ERP or PIM that owns the product master. That is where technical SEO and business systems meet. When an item is discontinued in the ERP, the store may delete the page outright, leave it live with no stock, or return a normal status with an empty template. Each choice sends a different signal. I agree a rule per case with you:

• Temporarily out of stock: keep the page, mark availability correctly, link to alternatives
• Permanently discontinued with a clear successor: redirect to it
• Retired with no successor: return a proper not-found or gone status

The same product data also feeds Google Merchant Center and price comparison services such as Prisjakt and PriceRunner. If the feed, the visible page and the structured data disagree on price or stock, trust suffers everywhere. My ecommerce SEO and ERP integration pages describe each side in more depth.

Encoded å, ä and ö: one page, one address

A browser shows a Swedish letter in the address bar, but underneath the URL is percent-encoded, and different systems do not always encode it the same way. Character normalization adds another trap: the same visible ö can be stored as a single character or as an o with a combining mark, which produces two distinct encoded strings. Add transliterated versions that a previous agency introduced, and a single category page can end up with several live addresses.

I trace how each address variant arises and where it is referenced:

• Internal links and navigation menus generated by the CMS
• Canonical tags, which must match the exact form you want indexed
• XML sitemaps, which must be valid and consistently encoded
• Hreflang references between Swedish and other language versions
• Redirect rules, which often match the decoded form but miss the encoded one
• Server logs, where crawler requests reveal which variants are still being fetched

Whether you keep Swedish letters in slugs or replace them is a decision for your team and brand; the technical requirement is that every system produces the same string. Once the rule is agreed, I write redirect logic for the stray variants and a test list your developers can run after each release. Domain names with Swedish letters exist too, but many businesses keep a plain-letter domain and avoid the extra handling.

Filters, sister sites and duplicate content across the Nordics

Duplicate content on Swedish sites usually has two sources. The first is faceted navigation. A clothing, furniture or spare-parts store with filters for size, color, material and brand can create far more URL combinations than products. Crawlers spend their time on those combinations while new items wait. I map which filter pages match real search demand, such as a brand within a category, and recommend that only those become indexable, with clean URLs, unique titles and internal links. The rest are kept out through consistent canonical, robots and linking rules, chosen to fit your platform's limits.

The second source is sister sites. Swedish companies often run versions for Finland, Norway or Denmark, sometimes on separate domains and sometimes sharing Swedish-language text with a Finnish version. Duplicate text across markets is not a penalty in itself, but it does need clear signals. Each version should point its canonical to itself and use hreflang to name its siblings; a canonical that points from the Finnish store to the Swedish one tells search engines to drop the Finnish page entirely.

Market choices, language codes and localization are covered on my international SEO page for Sweden. Here the focus is the implementation: templates, tags and redirects that behave correctly on every version.

Core Web Vitals, hosting and the weight of consent tools

Swedish visitors generally have good connections, which makes slow pages harder to excuse and easier to diagnose: the bottleneck is usually the page, not the network. Field data from Search Console and the Chrome UX Report shows which templates fail loading, responsiveness or layout stability for real users, and I trace each failure to its cause.

Common causes on Swedish sites include:

• Consent management scripts loaded early and blocking rendering, or a banner that becomes the largest element on the page
• Checkout, payment and chat widgets loaded on every page rather than only where they are used
• Tag managers that accumulate pixels and testing tools nobody owns any more
• Hero images served at desktop size to phones or lazy loaded when they should load first
• Distant hosting or missing caching, where a Nordic data center or a CDN edge close to Sweden would shorten server response

Consent is a legal matter as well as a technical one. What must wait for consent is a call for your legal advisor or data protection lead; I make sure the chosen setup loads efficiently and that the banner does not shift the layout. For the full method, see Core Web Vitals consulting. I do not promise that faster pages will raise rankings, but slow templates reliably cost conversions.

Replatforming, structured data and working with your agency

Swedish companies change webshop platforms, merge brands and redesign sites regularly, and each change risks the signals earlier pages earned. I build the redirect map from crawls, sitemaps, analytics and backlink data, check the new templates on staging, and verify redirects, canonicals and tracking on launch day. The SEO migration consulting page lists the full checklist.

Structured data deserves attention at the same time. Product markup should read price in kronor, availability and identifiers from the same source as the feed. Organization markup can carry your legal name and Swedish company registration number as an identifier, and LocalBusiness markup should match each Google Business Profile exactly. The schema markup page explains the model I use.

If you already have developers or a Stockholm, Gothenburg or Malmö agency, I do not replace them. I act as the independent technical reviewer who:

• Writes tickets in their tracker with acceptance criteria
• Reviews pull requests or staging builds for SEO side effects
• Checks Search Console and logs after each release

Everything happens remotely in English, with calls set in the Swedish working day. For keywords, local search and content, see the SEO consultant page for Sweden or the Sweden overview, or get in touch about an audit.