What technical SEO actually covers, and why it comes first

Technical SEO has a reputation for being the boring part, the work you do before the interesting work of writing and earning links. It is more accurate to think of it as the part that sets the ceiling. A page that cannot be crawled cannot be indexed. A page that is not indexed cannot rank, regardless of how good it is.

Crawling: can a search engine reach the page?

Search engines discover pages by following links and reading sitemaps. Pages that are orphaned, buried behind faceted navigation, blocked in robots.txt or reachable only through JavaScript interactions may never be found. Meanwhile, a site that generates thousands of near-duplicate URLs, filter combinations, session parameters, endless pagination, spends its crawl activity on pages that will never earn a position.

The practical questions are simple: which URLs exist, which of them should exist, and are the ones that matter reachable within a few clicks of the homepage?

Rendering: does the content exist when it is read?

Modern sites often assemble content in the browser. Search engines can execute JavaScript, but rendering is a second pass and it is not guaranteed to be complete or timely. Content that only appears after a client-side fetch, behind a tab interaction, or after infinite scroll is content you are asking a search engine to work harder to see.

Server rendering or static generation removes the question entirely. If the HTML response already contains the copy, the headings and the internal links, there is nothing to wait for.

A page that cannot be crawled cannot be indexed. Everything else is downstream of that.

Indexation: is the right version being stored?

Once a page is fetched, the search engine decides whether to keep it and which URL to treat as canonical. This is where duplicate content, inconsistent canonical tags, conflicting hreflang annotations and accidental noindex directives cause damage that is invisible from the front end.

• One canonical URL per piece of content, declared consistently and matching internal links.
• Trailing slashes, protocol and subdomain handled by redirects rather than left to chance.
• Parameters that do not change content excluded or canonicalised.
• hreflang reciprocal and pointing at pages that actually exist, where multiple languages are served.

Architecture: what does the structure say about importance?

Internal linking is the clearest signal you fully control. The pages you link to most often, from the most prominent places, are the ones you are telling search engines to treat as important. A flat, well-organised structure with descriptive anchor text does more for a priority page than most external link building.

This is also where cannibalisation gets resolved. When three pages target the same query, the fix is usually consolidation: one strong page, the others redirected or repurposed to address a genuinely different intent.

Performance and stability

Core Web Vitals measure loading, interaction responsiveness and visual stability. Their direct ranking weight is modest, but their effect on conversion is not, and a slow site is also more expensive to crawl. Image optimisation, restrained JavaScript, sensible caching and avoiding layout shifts cover most of the ground.

Where technical work stops

Technical SEO removes obstacles. It does not create demand. Once a site is crawlable, renderable, correctly indexed and sensibly structured, further technical refinement has diminishing returns, and the constraint moves to whether your pages answer a question better than the alternatives.

The order matters, though. Publishing excellent content onto a site with a broken indexation model is an expensive way to find out that the foundations needed attention first.