Technical SEO.
Technical SEO is the work on the site's infrastructure: what lets search engines and AI answers find, understand and index the content. The foundation that makes the rest of the SEO work count.
By Hans Sandblom · Senior SEO Specialist· Published · Updated Our view
Technical SEO is the foundation. If it doesn't hold, whatever you build on top matters less.
What technical SEO is.
Technical SEO is the part of search optimization that concerns how the site is built, not what is on it. If on-page SEO covers the content and off-page SEO the links pointing to it, technical SEO is what makes the page findable, readable and understandable by a search engine or an AI answer in the first place. Google's crawling and indexing documentation↗ describes the foundations.
Technical faults are also the most expensive kind. A site with weak content ranks poorly. A site that can't be indexed doesn't rank at all. The order matters: before more content is written or more links are built, the technical layer needs to be in place. Otherwise the later work falls away without effect.
For AI answers (AI Overview, ChatGPT, Perplexity) the bar has risen. An AI generating a live answer can't guess. It cites the source that's easiest to parse, fastest to fetch and most clearly structured. Technical SEO becomes an inclusion factor, not just a ranking factor.
Six areas within technical SEO.
The work breaks into six connected areas. They solve different problems but belong together: a fault in one has consequences in another.
Crawlability and indexing
robots.txt and sitemap tell search engines which pages to visit. Errors here mean Google misses half the site, or crawls pages you don't want indexed (filter URLs on e-commerce are a classic example).
Page speed and Core Web Vitals
LCP, CLS and INP measure loading, stability and interaction. Official ranking factor since 2021. Measurable via PageSpeed Insights, Search Console and CrUX. Web.dev Core Web Vitals↗ documents the thresholds. A slow site ranks lower and converts worse.
Mobile
Google indexes the mobile version by default (mobile-first). Missing content or functionality on mobile is invisible. Viewport meta, responsive images, touch-friendly elements.
Security (HTTPS/SSL)
HTTPS has been baseline for years. Mixed HTTP/HTTPS content, expired certificates or weak protocols hurt both trust and ranking.
Structured data
Schema.org↗ markup in JSON-LD tells Google and AI what the content means semantically: product, FAQ, article, event. The foundation for rich results and for being cited in AI answers.
Canonicalisation and URL structure
The same content on several URLs creates duplicates. Canonical tags point out the main version. URL structure (logical hierarchy, readable slugs, consistency) signals the site's architecture.
Crawl budget and large sites.
Crawl budget is the number of pages Google visits on your site within a given period. For smaller sites (below a few thousand URLs) it's rarely an issue, Google gets to everything. For larger sites, e-commerce with thousands of product variants or media sites with deep archives, the budget becomes a constraint.
Waste happens when crawl time goes to low-value content: filter URLs without canonicalisation, paginated series without crawlable links between pages, internal 404s, unnecessary redirect chains. What should be indexed then doesn't get there, or gets refreshed rarely.
The fixes are concrete. Block low-value URLs in robots.txt. Set noindex on taxonomies you don't want indexed. Clean up broken links. Keep a logical URL structure. Optimise XML sitemaps per content type. All measurable in Search Console under Crawl stats.
JavaScript rendering and SPAs.
More and more of the web is built as single-page applications (React, Vue, Angular) where content is rendered in the browser with JavaScript instead of arriving ready from the server. That creates a crawl problem: Googlebot has to render the page before it sees the content, and rendering is expensive. Pages land in a render queue and get indexed later, sometimes incompletely.
AI answers sharpen the problem. AI Overview, ChatGPT and Perplexity usually fetch the raw HTML without running JavaScript. Content that only exists after client-side rendering is invisible to them. What isn't in the page source can't be cited.
The fix is to deliver content rendered from the server already: server-side rendering (SSR), static site generation (SSG) or prerendering for crawlers. Critical content, headings and links should be in the HTML on the first response, not injected afterwards. We test with Search Console's URL inspection against the rendered HTML and compare raw source against the DOM.
Common misconceptions.
Technical SEO is often described in terms that sound either essential or impossible. Rarely true. Three recurring myths:
- 'Running PageSpeed Insights is enough.' No. PSI shows two things: a lab test (Lighthouse) and, when available, CrUX field data. It's the field data that counts for ranking, real visitors' experience over time which you also track in Search Console, not the lab value from a single test.
- 'Schema markup lifts ranking.' No, schema doesn't lift ranking directly. It raises the chance of rich results and of an AI citing your source. The effect shows in CTR and AI inclusion, not in top-ten position.
- 'A modern platform is enough.' The platform is a starting point, not a promise. Most issues we see on 'modern' sites are configuration faults: wrong canonical, missing hreflang, pagination without crawlable links, not platform faults themselves.
How we work with technical SEO.
We start with a current-state analysis. A full site crawl, Search Console data and comparison against your site's content. What's most worth fixing first differs between sites. Often the low-hanging fruit is crawl waste and mobile CWV, but the picture has to be drawn per case.
We deliver two things: a concrete action list prioritised by effect and effort, and implementation help. Either we code it ourselves (if you give platform access) or train your dev team. You decide the path that fits. Everything is documented so you can run the work forward without us.
We measure the effect in Search Console. What counts is what Google sees, not what a reporting tool claims. You own all the data from day one.
Get a free technical SEO audit.
Send us your URL. We'll run a technical review: crawl, Core Web Vitals, structured data, canonical. You'll get back a concrete action list.
Write to us →Frequently asked questions about technical SEO
What is technical SEO?
Work on the site's infrastructure so search engines and AI answers can find, understand and index the content. Six areas: crawlability, page speed, mobile, security, structured data and URL structure.
What is an example of technical SEO?
Fixing a wrong canonical tag so that several URLs stop competing for the same ranking. Or improving LCP (largest contentful paint) from three seconds to one, which lifts both ranking and conversions.
How do you run a technical SEO analysis?
A full site crawl combined with Search Console data gives a broad picture: indexing faults, CWV values, duplicate pages, broken links. The output is a prioritised action list by effect and effort.
How does technical SEO affect AI search?
AI Overview, ChatGPT and Perplexity cite sources that are easy to parse and clearly structured. Fast loading, clean HTML and structured data (JSON-LD) make the page a likely source. Technical SEO becomes an inclusion factor, not just a ranking factor.
Is technical SEO a one-off or ongoing?
Both. An audit catches the big faults. Ongoing monitoring catches drift: new CMS updates, expired certificates, CWV regressions. The cadence is set by the site's size and rate of change, not by a fixed template.