How to Do a Website SEO Audit, Step by Step

A technical SEO audit answers one question: can search engines find, crawl, understand and index the pages you want to rank, and is anything getting in the way? You do not need to be a developer to run one. You need a crawler, a checklist and a sensible order for fixing what you find.

This guide walks through that checklist step by step.

1. Crawl the site

Start by crawling your website the way a search engine would: begin at the home page and follow the links. A crawl gives you the list of pages that are actually reachable through links, which is often different from the list you think you have.

A good crawler:

• Follows robots.txt, so it sees the site as search engines are allowed to.
• Identifies itself with an honest user agent, so you can recognise it in your logs.
• Is polite: it respects Crawl-delay and does not hammer your server with parallel requests.

Pages that exist but are not linked from anywhere ("orphan pages") will not show up in a link-based crawl. Compare the crawl with your XML sitemap to find them.

2. Check status codes

Every URL returns an HTTP status code. Look for:

• 4xx errors (for example 404 Not Found): broken pages. Fix the link that points to them, restore the page, or redirect it to the closest replacement.
• 5xx errors: server problems. These can stop search engines from crawling and should be fixed first.
• Redirects (3xx): use a permanent redirect (301 or 308) when a page has moved for good. Avoid redirect chains (A to B to C) and loops; link straight to the final URL.

3. Check indexability

A page can return 200 OK and still be kept out of search results. For every page you want to rank, check that:

• It is not blocked by robots.txt.
• It does not carry a noindex directive (in a meta robots tag or an X-Robots-Tag header).
• Its canonical URL points to itself, or deliberately to the preferred version of the page.
• It is listed in your XML sitemap, and the sitemap lists only indexable, canonical URLs.

And the other way round: pages you do not want in search results, such as internal search results or thin tag pages, should be noindexed or not linked at all.

4. Check HTTPS

Every page should be served over HTTPS, and the HTTP version of each URL should redirect permanently to the HTTPS version. Watch out for:

• Mixed content: HTTPS pages loading images or scripts over HTTP.
• Two versions of the site (with and without www) that both return 200. Pick one and redirect the other.

5. Review on-page basics

These elements help search engines and people understand each page:

• Title tags: every page needs a unique, descriptive title. Search results usually show around 50 to 60 characters before truncating, so put the important words first. Titles that are much shorter often waste the space; titles that are much longer get cut off.
• Meta descriptions: not a ranking factor, but often used as the snippet in results. Write a unique summary for each important page. Missing descriptions leave the snippet to chance; very long ones are truncated.
• Headings: give each page one clear main heading (H1) that says what the page is about. Google has said that several H1s are not a problem for search, but a single clear H1 helps readers and assistive technology, so audits usually flag multiple H1s as a minor issue.
• Images: add descriptive alt text to meaningful images, and set their dimensions.
• Duplicate content: pages with identical titles, descriptions or body text compete with each other. Consolidate them or use canonicals.

6. Check internal links

Internal links tell search engines which pages matter and help them discover new ones. Look for:

• Important pages that are many clicks away from the home page.
• Links that point to redirects or broken pages.
• Generic anchor text ("click here") where a descriptive one would help.

7. Test speed and Core Web Vitals

Slow, unstable pages frustrate visitors. Test your key templates (home page, a category page, an article, a product page) on mobile and desktop. See Core Web Vitals explained for what the metrics mean and how to fix them.

8. Check structured data

If you use structured data (for example for articles, products, FAQs or breadcrumbs), make sure it is valid and matches what is visible on the page. Never mark up content that is not there, such as reviews you do not have.

9. Prioritise the fixes

An audit usually finds more issues than you can fix in one go. Work in order of impact:

1. Errors: anything that stops pages being crawled or indexed (server errors, broken important pages, accidental noindex, blocked sections).
2. Warnings: issues that weaken pages that are indexed (missing descriptions, redirect chains, duplicate titles).
3. Notices: small improvements (titles slightly too long, multiple H1s).

It also helps to look at what passed. Knowing which checks are already fine stops you from wasting time on them and shows progress over time.

10. Re-crawl and repeat

After you fix something, run the audit again to confirm it is resolved and that nothing new broke. Sites change with every release, plugin update and new page, so audit regularly: monthly is a good rhythm for most sites, and after every major release.

Run this audit in SEO Tracker Pro

The website audit in SEO Tracker Pro does steps 1 to 5 for you. Its crawler, SEOTrackerProBot, follows robots.txt and Crawl-delay, fetches one page at a time and only crawls the domain you added. The results are grouped into errors, warnings and notices, each with the pages it affects, plus the list of passed checks and an overall score. Audits never use credits: the Free plan includes one 25-page audit a month, and paid plans scan up to 5,000 pages with scheduled audits on Pro and Agency. See pricing for the limits.

Combine it with PageSpeed tests for step 7, and with rank tracking to see whether your fixes move your positions.