Technical SEO
Technical SEO is the work nobody sees and without which everything else falls flat. If search engines cannot find, read or classify your pages, the best copy in the world will not help.
This page explains what it usually comes down to in practice, for online shops as much as for large corporate websites.
Does this sound familiar?
New pages never make it into the index
Products, categories or landing pages are live, but Google's indexing ignores them. In Search Console they sit under discovered but not indexed.
The site is fast, the scores are bad
Everything looks fine in testing, yet the Core Web Vitals field data stays red. Lab and reality do not match.
You do not appear in AI answers
Rankings are stable, but you are not cited as a source in ChatGPT or Perplexity, even though competitors are.
Every tool says something different
Three analyses, three lists of issues. Nobody knows which are real and what should be tackled first.
Indexing: why pages never make it into Google
The most common reason for missing traffic is mundane: the page is not in the index at all. That question has to be settled before talking about rankings. In practice it almost always comes down to one of these four causes.
- Technical exclusion: robots.txt, noindex or a canonical pointing to a different page. Often a leftover from a staging environment.
- No internal links: pages that no other page links to are rarely found. They may sit in the sitemap, but that is no substitute for a real link.
- Crawl budget in the wrong place: filter, sorting and parameter URLs consume the attention your important pages are missing.
- Content that is too similar: Google treats the page as a duplicate and keeps only one variant. Typical with product variants and with several country versions in the same language.
Which cause applies cannot be guessed. Comparing server logs, sitemap and Search Console shows whether a page was requested at all, and if so, what happened next.
JavaScript rendering and AI crawlers
Google can execute JavaScript, but it does so with a delay and with no guarantee of completeness. On modern shop and website systems this creates a gap between what users see and what search engines store.
With AI systems the picture is clearer: AI crawlers generally do not execute JavaScript. If content is only assembled in the browser, it simply is not there for them. A page can rank in Google and still be completely absent from AI answers.
How to check this
Fetch the page source without JavaScript and compare it to the rendered view. If the main text is missing from the raw response, AI crawlers have nothing to read. Server logs additionally show which bots actually come by.
What helps
Server-side rendering or pre-rendering for the content that matters. This is no longer purely a performance measure but the precondition for appearing in AI answers at all.
How to build visibility in ChatGPT, AI Overviews and Perplexity on this technical foundation is shown on the GEO and AI visibility page.
Load time and Core Web Vitals
A PageSpeed score of 100 says less than most people think. It is a lab simulation under fixed conditions. What gets evaluated in the end is what real users experience, measured through Core Web Vitals in the field.
That is why I always start load time work with the field data and work backwards from there. The usual brakes are rarely exotic: render-blocking resources in the head, unoptimized images, too many external scripts and missing caching rules.
Clean measurement data is the foundation for this. If tracking and your data setup still need to be built, that is covered by the Go-to-Market Coaching. Technical SEO is one building block of several, all at a glance on the page about my SEO Services.
Structured data as a reading aid
Structured data tells search engines and AI systems explicitly what is on a page: a product with a price, a question with an answer, a company with a location. Without that markup the content has to be guessed.
In many projects it is missing entirely while competitors use it widely. Because the effect shows up quickly and the risk is low, it is usually the first technical measure I recommend.
Why tools alone are not enough
In larger projects several analysis tools often run in parallel and deliver contradictory results. One tool reports missing headings that are visibly present on the page. Another marks load time issues as resolved while measurement still shows clear delays.
Turning such lists into tickets unchecked keeps your team busy without changing anything. I verify every finding against the real page, filter out measurement errors and justify the remaining points well enough to earn a priority in the backlog.
Public checks to try out directly
SSR Text Checker
Compare what's in the initial HTML with the text that JavaScript generates in the browser. This shows you whether important content is missing for crawlers.
Open SSR Check →Online Presence Check
Check the visible presence of a domain and get a first assessment of rankings, traffic and competitive landscape.
View Online Presence →Keyword Ranking Check
Check a specific search query and get an initial assessment of how search engines rate the current ranking.
Open Ranking Check →AI Bot Log Analyzer
Upload your own server logs and check which AI crawlers actually visit your website and which URLs they request.
Open AI Log Check →Dashboard
Check the visibility, rankings, traffic and competitive landscape of your domain. After the analysis, you can embed the dashboard on your website with a personal embed code.
Embed Dashboard →Technical SEO by starting point
Frequently asked questions about technical SEO
Why is Google not indexing my page?
The most common causes are: the page is technically excluded, it cannot be reached through internal links, it is classified as too similar to another page, or crawl budget is used up by unimportant URLs. Comparing server logs, sitemap and Search Console shows which one applies.
How long does it take for a new page to be indexed?
On a well-linked website often a few days, on large or rarely updated sites several weeks. If it consistently takes longer, that is not a question of patience but a sign of a structural issue in internal linking or crawl control.
Is JavaScript bad for SEO?
Not in principle. Google does render JavaScript, but with a delay and with no guarantee of completeness. It becomes critical with AI crawlers: they generally do not execute JavaScript at all. Content that only appears in the browser does not exist for them. Server-side rendering or pre-rendering solves both at once.
Is technical SEO worth it if the content is already good?
Good content that search engines cannot access has no effect. Technology does not compete with content, it is the precondition for it. The reverse holds just as well: a technically perfect site without matching content still will not rank.
Is a PageSpeed score of 100 enough?
No. The score is a lab simulation. What counts is the field data from real users that Google collects in the Core Web Vitals. A page can score very well in the lab and still perform badly for real users on slow mobile connections.
Where do you start?
Technical SEO is rarely a single problem, it is a sequence. If you want to know which three points come first for you, the audit is the right entry point. If work is already running and what is missing is prioritization, ongoing support is the better fit.