Short answer

How do I do a DIY SEO audit?

Crawl the whole site, fix anything that stops pages being indexed, then broken links, then titles and descriptions, and re-check. That order is the whole audit. A free crawler finds the faults page by page, Google Search Console tells you which pages matter, and for a small or medium site the first pass fits in an afternoon.

Below are the five steps in order, what to look for at each one, how to tell a real fault from noise, and the situations where an audit you do yourself is not enough.

What a DIY SEO audit covers, and what it cannot replace

An SEO audit checks whether search engines can find, fetch and understand every page you want found, and whether each page says plainly what it is. Done yourself, it covers the technical and on-page side: indexing directives, status codes, links, redirects, titles, descriptions, headings and alt text. These are the faults that cost rankings for no reason, and they are all on your own site.

It does not cover what happens off your site: who links to you, what competitors rank for, and how many people search for queries you do not appear for yet. Nor can it judge whether your writing is good; it can only tell you whether search engines can see it.

Some jobs deserve an expert from the start, or at least an hour of one:

  • Moving domains or platforms. Every old address needs a redirect, and a careless migration can lose years of rankings quickly.
  • Very large sites. Tens of thousands of pages, faceted navigation or pages generated from a database make it an engineering problem.
  • A sudden fall in traffic. It has a cause, such as a manual action or a blocked section, and finding it fast is worth paying for.

Doing SEO on your own covers where the DIY approach runs out in more detail.

Step 1: Crawl your whole site for free

Work from what is actually published, not what you think is published. Enter your home page on the scan page. The crawler follows links from there, reads up to 5,000 pages, and lists every fault with the URL it is on. No account and no card. On a larger site you can ask for an email when the scan finishes.

Two options are worth knowing. Path filters and depth limits keep a scan to one section if that is all you want to check. And if your content only appears after JavaScript runs, turn on “Use browser”: the scan then compares the HTML your server sends with what a browser builds, which shows what a crawler that does not run scripts is missing.

While it runs, open Google Search Console, go to Performance, then Pages, set the range to three months and export it. You will use it in step 5.

Step 2: Fix indexing problems first

These faults are all-or-nothing: an excellent page that cannot be indexed earns nothing, so a single one of them outranks everything else in the audit. Check four things.

  1. Noindex. Any page you want found that carries noindex, in a meta robots tag or an X-Robots-Tag header. It often survives from a staging site or a plugin setting. The scan's noindex check lists every page that carries one.
  2. robots.txt. A Disallow rule blocking a section you want in search, or a Disallow: / left over from development that blocks the lot. The robots.txt tester shows how each rule resolves. Remember that robots.txt stops crawling, not indexing: to keep a page out of search, let it be crawled and use noindex.
  3. Canonicals. Each page's canonical should point to itself or to the version you want indexed. A canonical pointing at the home page, or at a URL that redirects or returns 404, is a fault.
  4. XML sitemap. It should list only URLs that answer 200, are indexable and are their own canonical, and it should be submitted in Search Console. The canonical checker and sitemap validator each check one address without a crawl; the rest are on the tools page.

Then confirm in Search Console's page indexing report, which lists pages excluded by a noindex tag or blocked by robots.txt. Anything there that you meant to be found goes to the top of your list.

Step 3: Broken links, redirect chains and orphan pages

  • Broken internal links. Links to pages that return 404 or 410. Either update the link at its source or redirect the dead address with a 301 to its closest replacement. One deleted page linked from fifteen others is one fix, not fifteen. The broken link checker lists every one across the site with the page it is on.
  • Redirect chains. A link that goes through two or more redirects before it lands. Point the link straight at the final address, so every internal link reaches a page that answers 200 directly.
  • Orphan pages. Pages in your sitemap or in Search Console that no page on the site links to. Add a link from a relevant page, or remove the page if it has no purpose.

Broken links to other sites matter less, but fix or remove them too: they are dead ends for your readers. The scan checks them after the crawl; the external link checker page explains how.

Step 4: Titles, meta descriptions, H1s and image alt text

  • Titles. Every page has one, no two are the same, and each says what the page is in the words your customers use, with the important words first. Missing or duplicated titles usually come from one template; the bulk meta tag checker lists them for the whole site.
  • Meta descriptions. They do not move rankings, but they shape the snippet people decide to click on. Write one or two sentences on what the page offers. Google may rewrite them; a good one is rewritten less often.
  • H1s. One clear main heading per page that matches what the title promises. The H1 checker lists pages with none or several.
  • Image alt text. Describe images that carry meaning; leave decorative images with empty alt text.

Do not tune titles to an exact character count or rewrite pages that get no impressions before fixing the ones that do. The 80/20 rule in SEO explains why.

Step 5: Prioritise the fixes and schedule a re-check

A scan here groups its findings by cause, so one title template or one dead destination is one task. Each task says what was observed, what to do, how many URLs it affects and how to verify the fix. Import the Search Console export from step 1 and the tasks are reordered by the clicks, impressions and click-through rate of the pages they touch. Severity still comes first, so a blocked page is never outranked by a cosmetic fault on a busy one.

Work through them in this order:

  1. Anything that stops a page you want found from being indexed.
  2. Faults on the pages with the most clicks and impressions.
  3. Template-level faults, which fix every page built from the template at once.
  4. Everything else, a few at a time.

Then re-scan to confirm the faults have gone. A free account keeps your scans together, compares each one with the last, and lets you schedule repeat checks. Search Console tells you, over the following weeks, whether the fixes changed what Google indexes and shows.

Questions people ask

How long does a DIY SEO audit take?
For a small or medium site, the first pass is an afternoon: the crawl runs while you open Search Console, the indexing checks take minutes, and the faults that are plainly mistakes can be fixed the same day. Rewriting titles and descriptions across many pages takes longer, unless they all come from one template, in which case it is one fix.
How often should I audit my own site?
Re-scan after anything structural: a redesign, a new theme or plugin, renamed or deleted pages, a platform move. Otherwise a full pass every month or quarter is enough for most small sites. With a free account you can schedule repeat checks, so faults that creep in are caught without you having to remember to look.
Is a free audit tool enough, or do I need a paid one?
For the technical and on-page checks in this audit, a free crawler and Google Search Console are enough. Paid tools add data about other people's sites: backlink indexes, competitor rankings and search volumes for queries you do not appear for yet. Doing SEO for free sets out what each free tool covers.
How is an audit different from an SEO checklist?
A checklist is the list of things to check; an audit is checking them against your actual site and fixing what you find. The DIY SEO checklist lists 25 checks. This audit runs the ones a crawler and Search Console can answer, in the order that pays off.

Start your audit with a crawl

Step 1 takes a minute to start and runs while you open Search Console. Free, no account, up to 5,000 pages.
Scan a site, free

Read next