Technical SEO: before you convince Google, the site has to be readable by Google

We have seen months of content work stuck behind one wrong line in robots.txt. The writing was good. The strategy was good. Google could not get through the door.

That is technical SEO: making sure the machine can arrive, read, and understand, before any fight over position.

The order of things

Picture a great store on an unpaved street. Its problem is not the storefront.

Every piece of semantic work on this site, from entities to AI citation, assumes the site is crawlable, fast, and readable. When that base fails, nothing else starts. This is called the technical foundation.

This silo gathers the tooling for checking that foundation. Every article follows the same house rule: a real screenshot, a dated figure, and the data owner’s caveat before the number.

The diagnosis Google hands you for free

Google itself tells you how it is reading your site, and almost nobody opens those screens.

Crawl stats show how much effort Google spent on your site and on what. The crawling article reads that screen on a real site, including the caveat that small sites rarely need it, and what showed up anyway. ¹

The AI audit crosses Search Console with a language model to find in minutes what used to take an afternoon. The method is published, prompt included.

And outside Google, Bing Webmaster Tools shows the traffic almost everyone ignores. The article opens with an on-screen notice that changes how you read the entire dashboard.

The tools of the people who audit

Crawl like a robot: Screaming Frog walks the site the way Google would. The article audits our own sister site and publishes the defect it found, on purpose, so you can reproduce the crawl as you read.

Audit page by page: Page Analytics lays Analytics data over the page itself. The article includes the table of public numbers for both extensions that share the name. An extension asking for access to your data gets checked before it gets installed.

Find what is missing: Keyword Gap crosses your site against your competitors and returns the list of what they all answer and you do not.

The rule that runs through the whole silo

A tool’s number is a model. The data owner’s number is a measurement. Both are useful, and they are not the same thing.

Semrush search volume, a crawler’s simulated position, a traffic estimate: all of it prioritizes work. Confirmation comes later, in Search Console and Analytics, where the data belongs to whoever operates the system. Every article in this silo declares which of the two types each number came from.

Where to start

Never audited anything? Start with the AI audit. One hour of work, and the map of problems appears.

Large site? Start with crawling, because crawl budget wasted on a large site is traffic that never gets born.

No idea what to write? Start with Keyword Gap, and walk away with a six-month plan.

Rather have us do it? Our consulting work opens with this entire diagnosis done for you.

Frequently asked questions

Does technical SEO solve things on its own?

No. It unblocks them. A technically perfect site with no answer to the intent is a paved street with no store. The foundation serves the semantic work, and never replaces it.

Do I need a paid tool?

Not to start. Search Console, Bing Webmaster Tools, and the free version of Screaming Frog cover the essentials. Each article states what its own version limits.

How often should I audit?

Crawling and errors, monthly. A full audit at every meaningful change: migration, redesign, restructuring. And any time traffic drops with no editorial explanation.

Notes and sources

1. Google, Search Central, Search Console crawl stats report. The verbatim quotes from the documentation, including the caveat about sites with fewer than a thousand pages, are in this section’s crawling article, with the date of access.