There’s an uncomfortable truth in web rankings: you can have the best content, the most careful design and the most refined keyword strategy — if your site has technical problems, Google simply won’t find it. Or worse: it will find it, but decide not to show it.
Technical SEO is that invisible layer that decides whether your site is crawlable, indexable and understandable for search engines. You can’t see it or feel it, but it decides everything else.
What exactly is technical SEO
Technical SEO is the practice of optimizing a site’s infrastructure so search engines can crawl it, index it and rank it correctly. While content SEO deals with what your site says, technical SEO deals with how it’s built.
Think of it as a building: content SEO is the decoration and the furniture. Technical SEO is the foundations, the electricity and the plumbing. If that fails, nobody’s going to stick around to admire the furniture.
To rank a site, Google needs three things: to be able to crawl it (its bots access the pages without blocks), to be able to index it (pages are stored correctly in its database) and to be able to understand it (the structure and code communicate what each page is about). Technical SEO takes care of all three.
Crawling and indexing: the front door
It all starts with the robots.txt file, the text file at the root of your site that tells Google what it can crawl and what it can’t. A common and costly mistake: accidentally blocking important pages (like services or the blog) with a misconfigured Disallow directive.
The next step is the XML sitemap: the list of all the URLs you want Google to know about. Without a sitemap, Google can take weeks to discover new pages; with one, it finds them in days. And indexing is the process by which Google stores a copy of your page in its database — if a page isn’t indexed, it can’t appear in any result, and that’s verified directly in the coverage report in Google Search Console.
Core Web Vitals: page experience as a ranking factor
Since 2021, Google has used Core Web Vitals as a ranking factor. They’re three metrics that measure the user’s real experience:
| Metric | What it measures | “Good” threshold |
|---|---|---|
| LCP (Largest Contentful Paint) | How long it takes for the page’s main element to load | 2.5 seconds or less |
| INP (Interaction to Next Paint) | How long the page takes to respond when the user interacts | Under 200 milliseconds |
| CLS (Cumulative Layout Shift) | How much the page shifts visually while loading | Under 0.1 |
Many sites fail these metrics without knowing it. The most common causes: uncompressed images or images without defined dimensions, too much JavaScript running in the background, third-party scripts (chats, analytics, pixels) competing for resources, and slow hosting that drags down everything else — we go deeper into both problems in My website is slow: what can I do?.
In addition, since 2019 Google has used the mobile version of your site as the main reference for ranking, even for searches made from a computer. It isn’t an optional detail: it’s the standard.
The technical errors that do the most damage (and almost nobody checks)
- Duplicate content. If several URLs show identical or very similar content, Google doesn’t know which one to rank and dilutes the authority between the two. It’s solved with a
canonicaltag that indicates which is the main URL. - 404 errors and misconfigured redirects. If you deleted a page with inbound links and didn’t redirect, that authority is lost. The rule: permanent removal goes with a 301, temporary with a 302 — and watch out for redirect chains, every extra hop loses value and slows down loading.
- Orphan pages. A URL with no internal link pointing to it. Google may find it through the sitemap, but without internal links it receives no authority and is considered low importance. It’s fixed with breadcrumbs and by linking from higher-weight pages.
- JavaScript that hides content. If content is generated dynamically with JavaScript, Google may not get to see it. The structural solution is to render on the server or serve HTML already generated at build time, so the text is available before the JavaScript runs — the approach we follow in our own projects, as we explain in WordPress vs. Astro: which technology suits your project.
- Missing structured data. Schema markup helps Google better understand your content and enables special features in results: review stars, prices, frequently asked questions. Without it, your result is generic even if the content is excellent.
The new horizon: GEO and visibility in AI engines
In 2026 there’s a shift that changes the rules of the game: a growing share of searches is already resolved with AI-generated answers. ChatGPT, Perplexity, Gemini and AI-generated summaries on Google are becoming an entry point as relevant as the traditional search engine — and to appear there, classic technical SEO is no longer enough on its own.
The fundamentals still matter (authority, content quality, structured data), but with new nuances: many sites unknowingly block AI crawlers (GPTBot, ClaudeBot, PerplexityBot) in their own robots.txt, and that leaves their AI visibility at zero. Format also weighs more than before — models cite a paragraph that answers a question directly more easily than one that rambles. We cover these and other strategies in depth in GEO: how to get ChatGPT to recommend your brand.
Technical SEO doesn’t show in the final design, but it’s the first thing we check in any audit — without a solid technical base, no content strategy performs as it should.
The good news is that much of the technical SEO work (speed, structure, structured data) also benefits AI visibility. The bad news is that it demands an additional layer of attention: which crawlers you let through and how you structure content so it can be cited.
It isn’t a layer you solve once and forget — it’s the foundation that holds everything else up, and it’s worth reviewing before investing in any other content strategy.