Technical SEO Basics, Explained Without the Jargon

Run any site through a free SEO audit tool and it will hand back a list of 40, 60, sometimes 100 "issues." Almost none of them matter as much as the report makes them look. Here's what technical SEO actually is, broken into the four things that decide whether a page can rank at all, and what's safe to ignore.

What does "technical SEO" actually mean?

Technical SEO is the infrastructure layer, not the words layer: it decides whether a search engine can reach, read, and trust a page at all, before content quality ever gets judged.

Most people learning SEO start with content: keywords, headings, how many words a page should have. That's real, but it's the second half of the job. Technical SEO is the first half, the plumbing underneath the content that determines whether a search engine can even access what you wrote. A perfectly researched, well-written page that a search engine can't crawl or can't index ranks for nothing, because it never enters the competition.

It's worth separating this from the newer AI-visibility layer too. What answer engine optimization actually means is a related but distinct discipline: formatting content so an AI tool can quote it directly. Technical SEO is the foundation both disciplines share. Allowing the relevant AI crawler matters for the same underlying reason, though not every AI system requires Google-style indexing first: some can still surface a blocked page's link and title through third-party discovery, even without reading its content directly.

Crawlability: can search engines even find your pages?

A search engine can only rank a page it can reach. Broken links, blocked crawler access, and orphaned pages, ones with no internal links pointing to them, are the most common reasons a page never gets crawled.

Crawling is the first step: a search engine's bot follows links from page to page, reading what it finds. Google's own documentation on crawling and indexing lays out the mechanics, how URL structure, sitemaps, and crawler rules like robots.txt all affect whether a bot reaches a page in the first place.

Three things break crawlability most often on real sites: a robots.txt rule blocking more than intended (a line meant to hide a staging environment can accidentally block the live site too, especially after a migration); orphaned pages, ones with no internal link pointing to them from anywhere else on the site, which are much harder for a crawler to find, sitemap or not; and broken internal links or redirect chains, where every dead link or multi-hop redirect burns part of a site's limited crawl attention, called crawl budget, a real factor on larger sites.

None of this requires guesswork. Check the robots.txt file directly, and an SEO crawler or a suitable CMS plugin can surface an orphaned-pages report if you go looking for it.

Indexing: getting found isn't the same as getting listed

Being crawlable doesn't guarantee a page gets indexed. Stray noindex tags, duplicate or near-duplicate content, and canonical-tag mistakes are the usual reasons a crawled page still doesn't appear in search results.

Crawling and indexing are two separate steps, and mixing them up is one of the most common technical SEO confusions. A search engine can crawl a page, read every word of it, and still choose not to add it to the index it serves results from. That choice usually comes down to a small number of signals: an accidental noindex meta tag left over from a staging build, a page judged too similar to another one on the site (thin or duplicate content), or a canonical tag pointing to a different URL as the "real" version.

The fix is usually simple: search site:yourdomain.com for the page, or check its status directly in a webmaster tool, to confirm whether it's indexed at all before assuming it's a ranking problem rather than an indexing one.

Site speed and Core Web Vitals: what the thresholds actually are

Google's three Core Web Vitals thresholds are specific and documented, assessed from field data at the 75th percentile, separately for mobile and desktop: Largest Contentful Paint under 2.5 seconds, Interaction to Next Paint under 200 milliseconds, Cumulative Layout Shift of 0.1 or under. Passing them is a floor, not a ranking guarantee on its own.

Google's Core Web Vitals documentation defines exactly three metrics that make up this standard, each covering a different part of the experience:

1. Largest Contentful Paint (LCP) measures loading speed: how long until the biggest visible element on the page has rendered. The "good" threshold is 2.5 seconds or under.
2. Interaction to Next Paint (INP) measures responsiveness: how long the page takes to visibly react after someone clicks, taps, or types. The "good" threshold is 200 milliseconds or under.
3. Cumulative Layout Shift (CLS) measures visual stability: how much content unexpectedly jumps around as a page loads. The "good" threshold is a score of 0.1 or under.

These are real, published numbers, not vague "make it fast" advice, which is what makes them useful. A page can fail all three and still technically be indexed and crawlable; it just competes at a real disadvantage against pages that pass, particularly on mobile, where slow connections make every millisecond count. The most common fixes are unglamorous: compress images, defer scripts that aren't needed immediately, and cut whatever is causing content to shift after it first appears.

Structured data: telling machines what your page is

Structured data (schema markup, usually written as JSON-LD) is a machine-readable label on top of the visible page, telling search engines and AI systems exactly what the content is, who wrote it, and how it's organized.

A headline, an author name, a publish date: obvious to a human glancing at a page layout, but none of it is obvious to a machine parsing raw HTML unless something states it explicitly. Structured data does that, in a standard vocabulary search engines and AI tools both read. Google's own structured-data documentation publishes a real case-study number behind the claim: Rotten Tomatoes added structured data to 100,000 pages and measured a 25% higher click-through rate on those pages compared to pages without it.

This isn't only a Google-ranking mechanism anymore. The same machine-readable labeling can help AI systems understand what a page is, though whether they quote it still depends on crawlability, relevance, and each system's own undisclosed process, a topic covered in more depth in the mechanism behind getting cited by ChatGPT. For most sites, the highest-value schema types are Article, Organization and Person; FAQPage stopped earning Google's FAQ rich result in May 2026, so use it mainly where other systems still benefit, matched exactly to what's visible on the page.

Sitemaps and robots.txt: the map and the rulebook

A sitemap tells a search engine which URLs exist. Robots.txt tells it which ones to leave alone. Neither one ranks a page by itself. A sitemap is optional, though helpful; an overly broad robots.txt rule can block crawling, but a noindex tag or access control, not robots.txt, is the reliable way to keep a page out of the index.

A sitemap is a file, usually XML, listing every URL a site wants search engines to know about. Google's own sitemap guidance is direct about who needs one most: large sites, sites with pages that aren't well linked internally, or new sites with few external links pointing to them, because a sitemap gives a search engine a direct list instead of relying entirely on discovering pages by following links. A small, well-linked site can survive without one; most sites are better off having one anyway, since it costs almost nothing to maintain.

Robots.txt does the opposite job: it tells crawlers what not to bother with, staging environments, internal search-result pages, admin areas. The most expensive mistake here isn't a missing robots.txt file. It's an overly broad rule in an existing one that blocks something it was never meant to touch.

What actually matters vs. audit-tool noise

Most SEO audit tools flag dozens of items with no real weight behind them. Crawlability and indexing decide whether a page can rank at all; Core Web Vitals can affect how well it competes and structured data can enable extra search features, but neither is a blocker on its own.

In priority order:

1. Crawl blockers. Anything stopping a search engine from reaching a page: robots.txt errors, broken navigation, orphaned pages. Fix these first; nothing else matters until a page can be reached.
2. Indexing errors. Stray noindex tags, canonical mistakes, and duplicate content that keeps a crawled page out of the index. Fix these second.
3. Core Web Vitals failures. Real, measurable performance problems that put a page at a genuine disadvantage. Fix these third.
4. Structured data gaps. Missing or broken schema on pages that should have it. Fix these fourth; it adds a meaningful edge, but a page still ranks and gets read without it.

Everything below that, meta description length, header-tag capitalization, minor alt-text phrasing, is the noise most audit tools present with the same red-flag urgency as a genuine crawl blocker. It isn't nothing, but it's the last 10% of the work, not the first. Spending a week fixing 60 cosmetic warnings before a real crawl blocker gets touched is the single most common way technical SEO effort gets wasted.

Where does this fit into a bigger SEO plan?

We're a UK marketing agency, founded in 2017, with 600+ clients since and 50+ active today, and technical SEO is the part of the job clients notice least and depend on most: nobody thanks you for a page that was always going to be findable, but everybody notices when a competitor's page outranks theirs for no visible reason.

Getting the four fundamentals right is the floor every other piece of SEO and AI-visibility work stands on. Content strategy, backlinks, and AEO formatting all assume a page a search engine can actually reach and read. Skip the floor and none of that other work gets a fair shot.

If you want an honest look at where your own site stands on these four, tell us about your business. We'll tell you straight which one is costing you the most, not sell you a 60-item punch list.