99creatix What technical SEO actually means, in plain words See our pricing
Home / Blog / SEO / What technical SEO actually means, in plain words
SEO · 8 min read

What technical SEO actually means, in plain words

Three questions a machine asks about your website, the jargon translated, and four checks you can run yourself in fifteen minutes without buying a tool.

The short answer

Technical SEO is three questions: can a machine reach the page, can it understand the page, and is this the right page to show. Every term in the field answers one of those three. On small business sites the damage is concentrated in a handful of failures: a blocking rule left over from the build, a redesign with no redirect map, duplicate homepages with no canonical, unresized images, thin near-identical pages, and no structured data. You can check the worst of these yourself in fifteen minutes with nothing but a browser.

Every price we charge is published. See the full list.

Technical SEO is three questions a machine asks about your website. Can I reach this page. Can I understand it. Is it the right one to show. Everything filed under the term is one of those three questions wearing specialist vocabulary.

Nearly everything written on this subject is aimed at people who do it for a living. That is reasonable, and it leaves business owners buying something they cannot evaluate. This piece is the owner's version: what the words mean, what breaks in practice, how to tell whether yours is broken without buying a tool, and what it is fair to pay.

If you read nothing else
  1. Technical SEO answers three questions: can a machine reach the page, understand it, and tell it is the right one. Every term in the field belongs to one of those three.
  2. The failures that matter most are total rather than partial. A page blocked from crawling does not rank badly, it is absent.
  3. Four checks, fifteen minutes, no tools: a site: search, reading your own robots.txt, loading the site on mobile data, and searching your own headline in quotation marks.
  4. Technical work makes a page reachable, readable and unambiguous. It does not make the page worth reading, and a flawless site saying nothing anyone searches for ranks for nothing.
  5. It is mostly a one-off correction plus light maintenance. An ongoing monthly technical retainer on a site that rarely changes is monitoring, which is fine if it is described as monitoring.

The three questions

Can I reach this page? A search engine finds pages by following links and reading a list you provide. If a page is linked from nowhere, blocked by a rule in a file most owners have never opened, or sitting behind a login, it might as well not exist. This is the category where the failures are total rather than partial: the page does not rank badly, it is absent.

Can I understand it? Once reached, the page has to be readable. That means the text is genuinely in the page rather than assembled afterwards by scripts that a machine may not run. It means headings that describe sections, images that carry a description, and a page that loads before anyone gives up.

Is it the right one? If four addresses show the same content, which is the real one? If you have a page for Manchester and one for Leeds and they are ninety per cent identical, which answers a search for either? Ambiguity here does not produce an error message. It produces a quiet decision made without you.

The vocabulary, translated

The terms you will see on a proposal, in plain words, with what actually goes wrong.

The termQuestionWhat it means, and what breaks
CrawlabilityReach Whether search engines can get to your pages. Breaks when a rule left over from when the site was being built is still blocking everything, which is more common than it sounds and usually dates from launch day.
IndexationReach Whether pages that were reached were actually kept. A page can be crawled and still not stored, usually because it was marked as not for indexing or judged to be a duplicate of something else.
XML sitemapReach A list of your pages handed over directly rather than left to be discovered. Cheap insurance. Breaks by going stale and listing pages that no longer exist.
Core Web VitalsUnderstand Three measurements of how the page feels while loading: how fast the main content appears, how quickly it responds when tapped, and whether things jump around as it settles. The third is the one customers actually complain about without knowing its name.
JavaScript renderingUnderstand Whether your words exist in the page or are assembled by scripts after it arrives. If the latter, a machine may see an empty shell. Most common on sites built with modern app frameworks rather than on ordinary business sites.
Schema, structured dataUnderstand Labels in the code stating plainly what the page is: a business, its address, an article, a price. It removes guesswork, and it is also what AI assistants read when they describe you.
Canonical tagsRight one A note saying which address is the real version when several show the same thing. Missing canonicals are why a site can compete against itself without anyone noticing.
RedirectsRight one Forwarding an old address to its replacement. The classic disaster is a redesign where nobody maps the old addresses, so every page that had earned anything returns an error and the rankings go with it.
hreflangRight one Tells a search engine which language or country version to show. Only relevant if you have more than one, and a frequent source of quiet errors when you do.

What actually breaks on small business websites

Audit reports run to eighty items. In practice a small number of failures account for most of the damage, and they are unglamorous.

  • A blocking rule left in place from the build. The site was hidden while it was being made and nobody removed the instruction afterwards. Total invisibility, trivial fix, and it can sit undiscovered for a year.
  • A redesign with no redirect map. Every old address returns an error. Whatever the previous site had earned is discarded, and the drop arrives a few weeks later when everyone has stopped watching.
  • Four versions of the same homepage at four addresses, none of them nominated as the real one.
  • Images straight from a camera. A single unresized photograph can outweigh the rest of the page combined. The most common cause of a slow small business site by a wide margin.
  • A page for every service and town, all near-identical. Thin pages built to catch searches, competing with each other and diluting the site.
  • Nothing in the code saying what the business is. No structured data, so both Google and the AI assistants are inferring from prose rather than reading a statement.

Checking yours without buying anything

Four checks, about fifteen minutes, no tools beyond a browser.

  1. Search for site:yourdomain.com in Google. The rough number of results is how many of your pages are stored. If it is far below your real page count, you have a reach problem. If it is far above, you probably have duplicates or thin generated pages.
  2. Visit yourdomain.com/robots.txt. If you see Disallow: / on its own, your site is asking not to be listed at all. That single line is the highest-stakes character string on most websites.
  3. Open your site on a phone, on mobile data, not office wifi. Count the seconds before you can read something. Over four and you are losing people before the content is even a factor.
  4. Type your best page's exact headline into Google in quotation marks. If your own page does not come back, it is not stored, and no amount of content work will help until that is fixed.

These are blunt instruments. They will not find a subtle canonical problem. They will find the failures that matter most, which are the total ones.

What technical SEO is not

It is not content, and it is not keywords. Technical work makes a page reachable, readable and unambiguous. It does not make the page worth reading. A technically flawless site saying nothing anybody searches for will rank for nothing, and that is not a technical failure.

It is also not permanent, but it is close. Most of it is a one-off correction plus light maintenance. Anyone selling you an ongoing monthly technical retainer for a site that does not change much is selling you monitoring, which is fine as long as it is described as monitoring.

One note on AI assistants, because the question always comes: the same work that lets a search engine read your pages is what lets an assistant read them too. Whether they can find and describe your business is a separate subject, and we cover it on the AI visibility page rather than here.

How we know this

The three-question framing, the grouping of terms under it, and the list of what most commonly breaks on small business sites are our editorial account drawn from client audits, not a ranked measurement or a published Google framework. The four self-checks are deliberately blunt and will not surface subtle canonical or hreflang problems. The $500 technical SEO foundation line, the $4,900 build total and the $1,350 a month SEO and AI search rate are our published prices, listed in full on the pricing page. The 47% organic traffic figure over 90 days is from our own client engagement and its measurement method is on the case study page.

What it should cost

On a new build with us it is a line called technical SEO foundation, $500, inside the $4,900 total, and it covers schema, speed, sitemaps and llms.txt from day one. Every line is itemised on the pricing page.

On an existing site it is a fix rather than a subscription. Ongoing SEO and AI search work starts at $1,350 a month, and technical maintenance is a small part of that; most of the money goes to content and the things that actually move rankings once the plumbing works.

For a UK fashion house we rebuilt the site in thirty days and cleared its technical SEO debt before touching content at all. Organic traffic rose 47% in 90 days while their paid spend came down. The case study sets out what was broken and how the number was measured. The sequence matters there: the technical work did not create the traffic, it removed the reason the existing pages could not earn any.

If you are holding a proposal full of the words in the table above, ask which of the three questions each line answers, and what specifically is broken on your site right now. A good answer is short and names actual pages. A vague one means the audit was generic.

Questions people ask

Short answers.

What is technical SEO in plain English?

It is the work that lets a search engine reach your pages, read them, and tell which page is the right one to show. Crawlability and indexation are about reaching. Speed, structured data and rendering are about reading. Canonicals, redirects and hreflang are about identifying the right version. Everything else in the field is a variation on one of those three.

How do I know if my website has technical SEO problems?

Four checks take about fifteen minutes. Search site:yourdomain.com to see roughly how many pages are stored. Visit yourdomain.com/robots.txt and look for a line reading Disallow followed by a single slash, which asks search engines to ignore the whole site. Load your site on a phone using mobile data and count the seconds. And search your best page headline in quotation marks to confirm the page exists in the index at all.

Is technical SEO a one-off or ongoing?

Mostly one-off, with light maintenance. The large corrections such as redirect mapping, canonicals and structured data are fixed once. What continues is watching for new breakage after changes to the site. If you are being sold an ongoing monthly technical retainer for a site that does not change much, ask what specifically is being done each month.

How much should technical SEO cost?

On a new build with us it is a $500 line inside the $4,900 total, covering schema, speed, sitemaps and llms.txt from day one. On an existing site it is a fix rather than a subscription. Our ongoing SEO and AI search work starts at $1,350 a month, and technical maintenance is a small part of that, with most of the money going to content.

Does technical SEO help with AI search too?

The same work that lets a search engine read your pages lets an AI assistant read them too, and structured data in particular states plainly what your business is rather than leaving it to be inferred. Whether assistants find and recommend you is a broader question involving reviews and third-party sources, which we cover separately on our AI visibility page.

The author

Who wrote this.

Written by Madiha Fatima CEO & Founder, 99creatix

Founded 99creatix in 2019 and has spent the years since building websites and running growth programmes for small businesses across the UK, EU, USA, Canada and UAE. Writes here about what things cost and why, on the theory that a client who understands the invoice is a better client to have.

This article
Published29 Aug 2026
Last reviewed29 Aug 2026
Reading time8 min
HubSEO
Related servicePricing
More from this hub

Keep reading.

Next step

Now you know the number.
Here is the whole list.

Every price we charge is published and itemised to the line. If it does not fit your budget, you will know in ten seconds instead of two weeks.