SEO & Marketing

The technical work starts with a simple question: which pages actually make it into the index. We read the status reports in Google Search Console, look for entries such as "Crawled, currently not indexed" or "Alternate page with proper canonical tag", run our own crawl and check whether sitemap, internal linking and indexed addresses tell the same story.

From there we move to what sits in the source code: canonical tags, robots.txt and meta robots, hreflang on multilingual sites, redirect chains left over from earlier relaunches, pagination and parameter URLs. WooCommerce and Shopware add their own traps, such as indexable filter and sorting URLs or product variants served as separate pages. WordPress adds automatically generated category and tag archives that rarely carry content of their own and still end up indexed.

On the content side we work with an architecture of topic pages and supporting detail pages that reference each other internally. Loading performance is judged on both field and lab data, meaning Core Web Vitals from real user sessions alongside Lighthouse runs. Priority follows what is blocking visibility, not what is quick to tick off.

Search optimization works where demand already shows up as search queries. For established products, services and spare parts that is usually the case. For a brand new category that nobody has a word for yet, nobody is searching either. In that situation ads, trade media and direct outreach do more than rankings.

Common triggers are a relaunch where existing positions have to be protected, a shop that so far sells almost exclusively through marketplaces, a site with plenty of visits but few inquiries, or a regional focus where local visibility is the point.

It makes less sense to start while the technical basis is missing. A site without server rendered content, without editorial upkeep, or with a structure that cannot be extended will slow down every measure taken. In those cases we begin with the technical layer or with the relaunch itself and build visibility on top of it. Visibility also never replaces the offer: if the service, the range or the messaging does not convince the market, better positions change little.

Answer engines such as Google AI Overviews, ChatGPT and Perplexity pull content from the open web, condense it and name only some of their sources. What gets cited is usually a self contained passage that answers a question directly, not a whole page.

In practice that means the question as a subheading, the answer in the first sentences below it, and the reasoning after that. Add structured data in JSON-LD that describes the organization, the offer, the location and frequently asked questions in machine readable form, plus company name, address and contact details that read identically everywhere.

Access decides the rest. Content that only appears once JavaScript has run in the browser is hard for many of these systems to reach. Blocking crawlers such as GPTBot or PerplexityBot in robots.txt means giving up mentions. Allowing them means releasing content for processing. We make that call together with you and document it. None of these systems promises that you will appear in an answer, and nobody can promise it on their behalf.