Short answer
Technical SEO is making sure Google can find, read and store your pages: crawling, rendering, indexing, speed and structure. Good content on a site Google can’t crawl is a book nobody can open. Roughly half of it you can check yourself, free.
Technical SEO is everything that decides whether a search engine can find your pages, load them, understand them and trust them. Not what the pages say — that is content. Whether they can be read at all.
I think it gets explained badly, usually as a 10,000-word checklist aimed at developers. Here’s the version for the person who owns the website.
What’s the difference between SEO and technical SEO?
Technical SEO is one part of SEO: the part that decides whether search engines can crawl, render and index your pages at all. The rest of SEO is what those pages say (on-page) and what other sites say about you (off-page), and neither can do anything if the technical part fails.
Technical SEO is the foundation. On-page SEO — the words, the headings, the structure of the argument — is the rooms. Off-page — links, mentions, reviews — is your reputation in the neighbourhood.
The metaphor earns its keep because it gets the dependency right: beautiful rooms on a cracked foundation eventually collapse, and a perfect foundation with empty rooms attracts nobody.
Where it breaks down is scale. A cracked foundation is catastrophic; most technical SEO problems aren’t.
They’re more like a house where one room has been wallpapered over the door. Everything else works fine.
That one room is invisible, and nobody notices until someone goes looking for it.
How a search engine actually reaches you
Four things have to happen, in order, before you can rank at all. Most technical problems are a failure at one specific step — which is why “my SEO is broken” is almost always a more specific problem than it sounds.
Crawl, render, index, rank
Discover
Google finds the URL — by following a link from a page it already knows, or from your sitemap. A page with no internal links pointing at it and no sitemap entry may simply never be found.
Orphan pages are more common than people expect, especially after a redesign.
Crawl
It requests the page. Anything blocking that request — a robots.txt rule, a login wall, a server that times out under load — stops everything downstream.
Render
It runs the page, including JavaScript, to see the finished result. If your content only appears after a script runs, this is where sites quietly lose. What Google renders is what Google judges.
Index
It decides whether the page is worth storing. Thin pages, near-duplicates and pages that contradict their own canonical tag get dropped here — and Search Console will tell you which, and why.
Rank
Only now does any of the content, keyword or link work matter. Everything above is the entry fee.
Figure
Most technical problems are a failure at one of these five steps
- 01Discover
- 02Crawl
- 03Render
- 04Index
- 05Rank
What “healthy” and “broken” actually look like
In short: Almost no real site is entirely one column. The job is finding which two rows are yours.
The parts, in the order they are worth fixing
Not a checklist of equals. This is roughly the order of impact I see in practice.
Fix first
Crawling and indexing. Can Google reach your pages and is it choosing to
keep them? Nothing else matters until this is yes. Blocked resources, accidental
noindex tags left over from a staging site, and canonical tags pointing at the
wrong URL are the usual culprits — and all three are invisible from the front end.
Duplicate URLs. The same page reachable at several addresses splits its own
signals. Trailing slash and non-trailing-slash versions, http and https,
www and bare domain, and endless URL parameters. I’d pick one, redirect the rest,
and say so with a canonical tag.
Core Web Vitals and real speed. Largest Contentful Paint, Interaction to Next Paint, Cumulative Layout Shift. As a ranking factor I’d call it modest and frequently oversold. As a reason visitors leave before reading anything, it’s enormous.
Important
Mobile. Google ranks based on the mobile version of your site, not the desktop one. If they differ, the mobile one is the one that counts.
HTTPS. Table stakes now. A browser warning costs you more visitors than the ranking signal ever cost you positions.
Site architecture. How pages link to each other tells Google what you think matters. Anything important should be two or three clicks from the homepage, and should have real internal links pointing at it from related pages.
Worth having once the above is done
Structured data. Schema markup doesn’t make you rank higher. It makes you eligible for rich results — stars, FAQ dropdowns, prices — which lifts how many people click you at the position you already hold.
Sitemaps and robots.txt. A sitemap lists what exists; robots.txt says what to skip. Neither is glamorous and both are quietly load-bearing, particularly on large sites.
What do canonical tags and robots directives do?
A canonical tag tells Google which URL is the main version when several addresses show the same page. Robots directives control access: robots.txt tells crawlers which URLs not to request, and a noindex tag tells Google not to show a page in search results.
Canonical tags are signals, not commands. Google ranks redirects and rel="canonical" as a “strong signal” and a sitemap entry as a “weak signal”, and says they work better combined (Google, consolidating duplicate URLs).
That’s why the duplicate URL fix above is a redirect plus a canonical, not a canonical alone.
The robots mistake I see most is using robots.txt to hide a page. Google’s own documentation says robots.txt “is not a mechanism for keeping a web page out of Google”, and a blocked URL can still be indexed if other sites link to it (Google, robots.txt introduction).
The tool for that job is noindex, and noindex only works if Google can crawl the page: block it in robots.txt and the crawler never sees the noindex rule. Use robots.txt to stop crawling of things like internal search results, and noindex to keep a page out of the index. Never both on the same URL.
If a page you do want in Google is missing, my guide to why a website isn’t showing up on Google works through the causes in order.
Does technical SEO matter for AI search?
Yes. AI assistants that search the web use their own crawlers, and your robots.txt decides which of them can read your pages. The key split is between training crawlers and search crawlers: blocking a training crawler keeps your content out of model training, while blocking a search crawler keeps you out of that assistant’s answers.
OpenAI documents both kinds (OpenAI, overview of OpenAI crawlers). Disallowing GPTBot “indicates a site’s content should not be used in training generative AI foundation models”, while sites that block OAI-SearchBot “will not be shown in ChatGPT search answers”.
Google’s training control is Google-Extended, a robots.txt token rather than a separate crawler. It governs whether content may be used to train Gemini models, and Google states it “does not impact a site’s inclusion in Google Search nor is it used as a ranking signal” (Google, common crawlers). You can opt out of training without touching your Google rankings.
Everything else is ordinary technical SEO. A page that times out, or only shows its content after JavaScript runs, is a risk with any crawler. Check your robots.txt for a blanket Disallow you didn’t intend.
You’ll also see llms.txt recommended. It is a proposal for a file at your site root that summarises your content for AI agents, not an agreed standard. It’s cheap to add, but I’d treat it as optional, not as a fix.
How do you find and fix broken links, 404s and redirect chains?
Crawl the site every few months and fix problems at the source: point internal links straight at the live URL, not at a 404 or through a chain of redirects. A 404 isn’t a penalty, but Google removes indexed URLs that start returning one, so a 404 on a page that used to rank loses that page’s traffic.
Google’s documentation says URLs returning a 4xx status aren’t indexed, and already-indexed ones “are removed from the index” (Google, HTTP status codes). The damage is in what the 404 replaced: a page that ranked, or that other sites linked to, now passes nothing on.
For those, a 301 redirect to the closest live equivalent keeps the visitor and the link. For URLs that never mattered, a 404 is the correct answer.
Redirect chains (A redirects to B, which redirects to C) build up with every redesign. Google’s crawlers follow up to 10 redirect hops by default, so a chain rarely stops crawling outright, but each hop slows the page and adds one more rule to break at the next migration.
Update internal links to the final URL and collapse each chain into a single redirect.
The routine after any redesign, URL change or new plugin: run a crawl, sort by status code, fix every internal link pointing at a 3xx or 4xx URL, then check the Search Console Pages report for “Not found (404)”.
What’s on a technical SEO checklist you can run yourself?
A basic technical SEO checklist has six checks you can do free in an afternoon: the Search Console Pages report, PageSpeed Insights on three pages, your site on a phone, HTTPS and domain variants, a site: search, and a crawl for broken links and redirects.
None of this needs a developer. Work down the list — it will tell you whether you have a problem worth paying someone to look at.
The free self-audit
0/ 6
That is the whole list. Anything still unticked is your next job.
Free tool · Core Web Vitals reader
Is your page fast enough by Google’s own standard?
Run a real page (not just the homepage) through PageSpeed Insights and copy the three numbers from “Discover what your real users are experiencing”.
- Enter any of the three numbers.
Thresholds from Google’s web.dev: good is LCP ≤ 2.5 s, INP ≤ 200 ms, CLS ≤ 0.1; poor is above 4 s, 500 ms and 0.25. Measured at the 75th percentile of real visits.
What tools do you need for technical SEO?
Three free tools cover most technical SEO for a small site: Google Search Console for indexing, PageSpeed Insights for speed and Core Web Vitals, and a crawler such as Screaming Frog, which is free for up to 500 URLs per crawl. Paid SEO platforms mostly add scale, scheduling and reporting.
| Tool | What it tells you | Cost |
|---|---|---|
| Google Search Console | Which pages Google has indexed, why others are excluded, and which searches you appear for | Free |
| PageSpeed Insights | Speed and Core Web Vitals for a single URL | Free |
| Screaming Frog SEO Spider | Every URL a crawler can reach, with status codes, redirects, titles, canonicals and noindex tags | Free up to 500 URLs per crawl; a paid licence removes the limit |
Search Console is the only one of the three that shows what Google has actually decided about your site. The other two show what Google would see if it looked.
A crawler is where broken links, redirect chains and stray noindex tags turn up, and the free version stops at 500 URLs, so a larger site needs the paid licence or a crawl done in sections.
How complicated is technical SEO?
About half of technical SEO needs no technical skill: reading Search Console, running PageSpeed Insights, testing on a phone and checking HTTPS. The other half, such as JavaScript rendering, structured data and site migrations, needs someone who has done it before, because getting it wrong can lose rankings you already have.
Being honest about the line: JavaScript rendering problems, structured data implementation, restructuring a site’s architecture without breaking what already ranks, untangling redirect chains across a migration, hreflang for multiple languages, diagnosing why Core Web Vitals fail on real devices when the lab score looks fine, and recovering from a manual action.
Those need someone who has done them before, mostly because the failure mode of getting them wrong is worse than not doing them at all.
What is a technical SEO audit?
A technical SEO audit is a structured check of everything in this guide on one site: crawling, indexing, speed, mobile, duplicates, redirects and structured data, ending in a prioritised list of what to fix. The checklist above is the free version; a full audit adds a complete crawl, Search Console analysis and the fixes that need a developer.
Mine comes as a written report for £399 ($529), or £699 ($929) with the fixes done, and it’s free as part of a retainer. The SEO audit page sets out what’s covered and what you get at the end.
The short version
Technical SEO isn’t a competitor to content — it’s the thing that decides whether the content is allowed to compete. In my view most sites need a few specific problems found and fixed once, then a check every few months to catch drift.
If you aren’t sure whether yours has anything wrong with it, the Search Console Pages report is free and takes five minutes. If that turns up something you would rather not untangle yourself, that’s the job I do.
Common questions
What is technical SEO in simple terms?
Everything that affects whether a search engine can find your pages, load them, understand them and trust them — as opposed to what those pages actually say. Speed, crawlability, mobile, security, structure and structured data. It's the foundation the content sits on.
Is technical SEO a one-off job or ongoing?
Mostly one-off, then occasional. The big structural fixes are done once and stay done. What needs watching is drift — a redesign that breaks internal links, a plugin that slows the site down, a page template that quietly stops being indexed. A check every few months catches all of it.
How do I know if my site has technical problems?
Google Search Console tells you free. The Pages report shows what's indexed and what's excluded, and the reason for each exclusion. If pages you care about are sitting under 'Crawled — currently not indexed' or 'Discovered — currently not indexed', you have a technical problem worth investigating.
Does site speed really affect rankings?
It's a ranking factor, but a modest one, and it's routinely oversold. Speed matters far more for whether visitors stay than for where you rank. A page that takes six seconds loses people before it ever gets the chance to be judged.
Do I need structured data?
It won't make you rank higher on its own. It makes you eligible for rich results — the star ratings, FAQ dropdowns and price labels in search results — which raises how many people click at the same position. That’s worth having, but fix crawling and indexing first.
Want it done on your site?
Send me the address. The free audit is a short, plain-English read on what to fix first.
Get a free SEO audit