Technical SEO · 8 min read

A 30-minute technical SEO audit you can run yourself

Before you write another word of content, check whether Google can actually see the pages you already have. These are the checks I run first, in this order, on every site I pick up.

By Sumanjali Prasanna Runku · 12 May 2026

Technical SEO has a reputation for being complicated. Most of the time it is not. The expensive problems are usually boring: pages blocked by a stray directive, two versions of the same URL competing with each other, or a sitemap that has not been updated since the last redesign.

You can find the majority of them in half an hour with Search Console, a browser and a spreadsheet. Here is the exact order I use.

1. Is the page indexed at all?

Start with the URL Inspection tool in Search Console for the page you care about most. It tells you whether the URL is on Google, when it was last crawled, and — most usefully — why it was not indexed if that is the answer.

Then check the Pages report. Separate the two questions that people often merge into one:

• “Discovered — currently not indexed” usually means Google knows about the URL but has decided it is not worth crawling yet. Thin or duplicated pages end up here.
• “Crawled — currently not indexed” often means the page was fetched and judged low value. That is a content and duplication conversation, not a technical one.

Write down every important URL that is not indexed. That list is your audit.

2. Check robots.txt and noindex before anything else

Two lines of configuration can hide an entire site, and both are easy to add by accident during a staging launch.

• Open /robots.txt and read it line by line. Look for a Disallow: / that should have been removed.
• View the page source and search for noindex. Also check the X-Robots-Tag response header, which can apply the same instruction without appearing in the HTML.

I have seen a site lose months of progress because a “discourage search engines” checkbox stayed ticked after go-live. It takes ten seconds to rule out.

3. Canonicals: one page, one address

Every page should declare one canonical URL, and that URL should be the version you actually want ranking. Common failures:

• The canonical points to a different page by mistake, often a leftover template value.
• Both http and https, or both www and non-www, resolve and are treated as separate pages.
• Tracking parameters create near-duplicates that each collect a few impressions and none of the authority.

Pick one host and one protocol, redirect the others permanently, and be consistent in internal links.

4. Redirects and broken links

Crawl your own site with a free tool, or click through the main navigation and footer yourself. You are looking for three things: redirect chains (A → B → C), redirect loops, and links that return a 404.

Chains waste crawl budget and dilute signals. Flatten them so every old URL points directly at the final destination. Fix internal 404s, because they are the ones you control.

5. Sitemaps and internal linking

Your XML sitemap should contain only indexable, canonical URLs — no redirects, no 404s, no noindex pages. If the sitemap is generated automatically, check that it excludes category archives and tag pages you do not want indexed.

Then compare what is in the sitemap with what is actually linked from the site. Pages that appear in one but not the other are usually orphans, and orphans rarely rank.

6. Titles and headings are technical too

This is where technical and on-page overlap. A template that appends the brand twice, or a CMS field that falls back to the post ID, produces hundreds of weak titles at once.

• Check for duplicate titles across the site.
• Check for missing or duplicated H1s.
• Check that the title describes the page, not the company slogan.

One template fix can improve more pages than a month of individual edits.

7. Speed, mobile and Core Web Vitals

You are not chasing a perfect score. You are looking for something that would genuinely annoy a visitor: a hero video that downloads 13 MB before the page is readable, fonts that block rendering, images far larger than their container.

Test on a throttled mobile connection, not your office wifi. If the page takes more than a few seconds to show useful content, that is the finding — not the number at the top of the report.

Turning the audit into a plan

Put every finding in a sheet with two columns: impact and effort. Then work in this order:

1. Anything blocking indexing, today.
2. Duplicate and canonical problems, because they affect whole sections at once.
3. Template-level title and heading fixes.
4. Redirect chains and internal 404s.
5. Speed work, starting with the heaviest asset.

Re-check the same four reports after two to four weeks. If nothing moved, the fix was not the bottleneck — and that is useful information too.

Want a second pair of eyes? I run free audits and send back a prioritised list with the quick wins marked. Request one here.