What is technical SEO, and what does it include?
Technical SEO is the part of a site that decides whether Google can reach the pages at all. I explain it as plumbing: nobody admires it, and nobody thinks about it until a pipe bursts. It's also the first thing I open in an audit that ranks every fault by the clicks it costs.
Google describes Search in three stages. Crawling is its programs downloading your pages. Indexing is Google reading those pages and storing them in a large database. Serving is picking results from that database when someone searches. Google says it doesn't guarantee it'll crawl, index or serve any page, so the job isn't forcing anything into Google. It's taking away every reason for Google to say no.
| Stage | What I tell the owner | What usually breaks it |
|---|---|---|
| Finding | Google has to know the page exists, mostly by following links from pages it already knows | A page nothing links to, or a sitemap that lists the wrong addresses |
| Crawling | Google's crawler has to be let in and get a normal answer from the server | A robots.txt block, a failing server, a login wall |
| Rendering | Google loads the page like a browser, scripts and all, to see what a visitor sees | Text or links that only appear after a script Google can't run or reach |
| Indexing | Google decides the page is worth storing, under one address | A noindex tag, a copy at another URL, a page with little on it |
| Serving | Google shows the stored page for a search it matches | This is where content and links take over from the technical side |
How is it different from on-page and off-page SEO?
I use the building again. On-page SEO is what's inside each room: the words, the headings, the titles people see in Google. Off-page SEO is the building's reputation: other sites linking to it and talking about it. Technical SEO is the pipes, the wiring and the front door. You can furnish a room beautifully, and if the door's locked, nobody's walking in.
| Kind | The question it answers | Typical work |
|---|---|---|
| Technical | Can Google reach, read and store this page? | Indexing settings, redirects, canonicals, sitemaps, speed, rendering |
| On-page | Does this page answer the search it's meant for? | Titles, headings, copy, internal links, image alt text |
| Off-page | Do other sites trust it enough to point at it? | Links and mentions from other sites |
The three overlap. Internal links are page copy and plumbing at the same time, since they're how Google finds most pages: Google's SEO starter guide says it primarily finds pages through links from pages it already crawled. Speed sits on the technical side too, but a visitor feels it as a page that's slow to use.
The order isn't negotiable, though. If Google can't index a page, no rewrite and no link will make it show up, so I check the technical layer before anyone spends money on content.
How I explain it on a first call, in plain English
This is the order I take an owner through, over a shared screen. I teach technical SEO basics on the owner's own site, not on a diagram of crawlers, because a page they know makes every step stick.
- Step 01
Start with the plumbing
One sentence on what technical SEO is, and one on what it isn't: writing or link building. Nobody should think they're being sold three things at once.
- Step 02
Open their Search Console
Search Console is Google's own report on the site, so it's the fairest place to start. The Page indexing report shows how many URLs Google has indexed and a reason for each group it hasn't. Google also says not to expect every URL to be indexed, which settles a lot of worry early.
- Step 03
Pick one page they care about
A site-wide score means little to an owner. Their best service page means everything. I run it through the URL Inspection tool, which Google names as the way to check how it sees a page, and read the result out in plain words: indexed or not, which address Google chose, and when it last crawled it.
- Step 04
Turn each fault into what it costs
A noindex tag isn't interesting on its own. A service page that can't appear in Google is. I tie every fault to a page and, where Search Console has them, to its clicks.
- Step 05
Say what I don't know yet
Some faults only show up in a full crawl, and Google's starter guide says some changes take effect in a few hours while others could take several months. So I don't give an owner a date.
Rather have me do this? One person, whole job.
Book a call on UpworkHow I check the parts of technical SEO on every site
These are the parts I go through on every WordPress site, each with the plain English version I give the owner. The order doesn't change with the size of the site.
- i.
Can Google get in, and is anything shut out?
Robots.txt is the note on the front door telling crawlers where they can't go, and one stray Disallow can shut a whole folder out of the crawl. A noindex tag is a different request, to keep a page out of results. WordPress has a box called Discourage search engines from indexing this site, and when it's ticked, every page carries that request.
- ii.
Does every page have one address?
The same page at www and non-www, with and without a trailing slash, or with tracking tags on the end, is several doors into one room. Google says it chooses a single canonical URL per piece of content, and I check it's choosing the one the owner wants. A site with French and English versions of a page also needs hreflang, which Google describes as the way "to tell Google about the variations of your content."
- iii.
Do old addresses still lead somewhere?
A page that moved needs a permanent 301 redirect to its new home. I list every old URL that returns a 404 and every redirect that hops more than once.
- iv.
How does Google find the pages that matter?
Mostly through links, so I map which pages get them and which are buried. The sitemap's the backup list. The starter guide says one isn't required, but a sitemap full of redirected or noindexed pages contradicts the site, so I compare it with what's actually live.
- v.
What does Google see after the scripts run?
Google says it runs the JavaScript it finds with a recent version of Chrome. I compare the page source with the rendered page, and if the main text, the menu or the links only appear once a script runs, I check that Google's copy has them too.
- vi.
Is it quick for real visitors?
Core Web Vitals are Google's three measures of loading, responsiveness and visual stability. Its targets are LCP within 2.5 seconds, INP of 200 milliseconds or less and CLS of 0.1 or less, and I read the field data real visitors produce, not one lab test.
- vii.
Is every page on HTTPS?
Google's page experience checklist asks "Are your pages served in a secure fashion?" I check that every old address, the http one and the www one included, lands on the HTTPS page in a single redirect.
- viii.
Does it work on a phone?
Google says it uses the mobile version of a site's content, crawled with the smartphone agent, for indexing and ranking. So I check the phone version carries the same text, links and structured data as the desktop one, not a trimmed copy.
- ix.
Is the structured data valid?
Structured data tells Google what a page is: an article, a product, a business. I test it in the Rich Results Test, remove types the visible page doesn't support, and keep one plugin in charge of it, because two plugins printing markup is a common WordPress fault.
- x.
Can AI crawlers reach it?
AI tools send their own crawlers, and robots.txt can name each one. OpenAI's documentation says a site can allow OAI-SearchBot to appear in ChatGPT search while disallowing GPTBot for model training, and Google says its Google-Extended token doesn't affect inclusion in Google Search. I check the rules say what the owner actually intends.
How do I know if my site has a technical problem?
Owners notice the symptom, not the cause. These are the ones worth taking seriously.
A page you know exists isn't in Google. Check it in the URL Inspection tool. If it says the page is crawled but not indexed, or discovered and never crawled, those two statuses have their own diagnosis.
The not indexed side of the Page indexing report keeps growing. A few reasons there are normal. A reason that jumps by hundreds of pages in a week means a plugin, a theme or a setting changed. I've written up why a website isn't showing up on Google as a separate walk-through.
Clicks dropped after a change. A redesign, a host move or a permalink change all touch the technical layer. If the drop starts that week, that's where I look first.
Strange pages turn up in Google. On one UK client site, a location page template went live by accident, indexable, with a placeholder where its title should've been. Only a crawl of every published URL finds pages like that, because nobody links to them on purpose.
A plugin's green light isn't a technical check. SEO plugins score the page you're editing against a keyword. They don't tell you whether Google indexed it, which address it chose, or what it saw after rendering.
Who should fix it: the owner, a developer or a consultant?
It depends on where the fault lives. Some fixes are a checkbox, some are code, and some need someone to read Search Console and decide the fix first.
| Fault | Who fixes it | Why |
|---|---|---|
| Discourage search engines ticked on a live site | The owner, in two clicks | It's one box under Settings, then Reading |
| Noindex on the wrong pages, wrong canonicals, redirect gaps | Me | They're SEO settings and redirect rules, and the fix depends on the diagnosis behind it |
| Reasons piling up in the Page indexing report | Me | Each reason traces to a setting, a plugin or a template, and the tracing is the work |
| A slow theme, a broken template, server errors | A developer, working from my findings | It's build work, and I don't take build work on |
| Thin or duplicate pages | Me, with the owner signing off the words | Merging or pruning pages is an SEO call, but it's their business the words describe |
When the damage is concentrated in indexing, the follow-on job is getting the right pages crawled and indexed, and the audit tells you whether you need it. If you'd like to know who you'd be dealing with first, here's how I work. It's one person on every job, and every call is remote, over Zoom, Meet or async notes.
If you'd rather have one person tell you what's wrong in plain English and then fix the SEO side of it, that's the audit I run. Your developer keeps the build.
Book a call on Upwork
