The Technical SEO Checklist Every Site Should Pass Before Asking An Agency

Technical SEO is a handicap you either carry or remove. It rarely wins you a position on its own, and it will absolutely stop good content from earning one.

These are the checks worth running before anyone writes another page. Most are free. Most sites fail several.

Indexation

Confirm the pages you care about are actually indexed. Not submitted, indexed. The gap between the two is where a surprising amount of wasted content sits.

Check for pages blocked in robots.txt that should not be, noindex tags left over from staging, and canonical tags pointing at the wrong URL. A canonical pointing to the homepage across a whole template is a classic, and it quietly removes every page in that template from consideration.

One URL per page

Trailing slash and non-trailing slash. With and without www. HTTP and HTTPS. Uppercase and lowercase. Query parameters that produce the same content.

Each variation that resolves independently splits the signals for that page. Pick one form, redirect the rest with a permanent redirect, and make every internal link use the chosen form.

The sitemap matches reality

Every URL in the sitemap returns a 200. Nothing in it redirects, 404s or is blocked.

A sitemap full of dead URLs is a trust problem. It tells the crawler your declarations are unreliable, and it wastes the crawl budget you want spent on the pages that matter.

Server response time

Time to first byte is the ceiling on every other speed metric. If the server takes eight hundred milliseconds to start responding, no amount of image optimisation fixes the page.

Measure it first. It is the one number that constrains all the others, and it is usually a hosting or caching problem rather than a front-end one.

The largest element loads first

The biggest visible element in the first screen, usually a hero image or a headline block, should not be lazy loaded, should not wait on JavaScript, and should be preloaded.

Lazy loading a hero image is the single most common self-inflicted speed wound. The attribute exists to defer things below the fold, and applying it site-wide delays exactly the element being measured.

Layout stability

Content that shifts while loading is measured and it damages both ranking and conversion.

The causes are consistent: images without dimensions, fonts that swap and reflow the text, and injected banners that push content down. All three are fixable in a day.

Mobile is the version that counts

Indexing is based on the mobile rendering. If content, links or schema are present on desktop and absent on mobile, they effectively do not exist.

Check the mobile HTML rather than the mobile appearance.

HTML that survives without JavaScript

Content rendered only after client-side JavaScript executes is crawled inconsistently, and answer engines are worse at it than search crawlers are.

View the page source, not the inspector. If the main content is not in the source, it is at risk.

Redirect chains

Every hop costs time and dilutes signals. Redirects should go straight to the final destination, one hop, no loops.

Chains accumulate silently across a few years of site changes, and nobody notices until someone runs a crawl.

Structured data that validates

Broken schema is worse than no schema. Run the pages through a validator and fix the errors rather than adding more types.

Orphan pages

Any page with zero inbound internal links. On most sites this is a longer list than anyone expects, and it usually includes commercial pages someone paid to have written.

An llms.txt file, with the right expectation

It costs nothing to publish and it is not a ranking mechanism. Treat it as documentation of your entity for the engines that choose to read it, not as a lever.

How to work the list

Fix the ones that block indexation first, because nothing else matters until the page can be seen. Then speed. Then structure.

Then stop, and go and write something worth ranking.