The 42-point SEO audit checklist, in the order we work through it

An SEO audit checklist earns its keep when every item has a pass condition and the items are worked in the right order. Start with indexation. Search Console’s Page indexing report shows which of your pages Google has indexed. Google says you should expect only your canonical pages to be indexed, not every URL on the site (Google Search Console Help). A page outside the index cannot rank, so every other check waits on this one.

Below are the 42 checks we work through when we open an account, in that order, each with the condition that counts as a pass and where to find the answer. You can do the whole first section without a developer.

The short version

• Work in this order: indexation, duplicate URLs, redirects, internal links, on-page, then speed. A page Google has not indexed cannot rank, so later fixes do nothing for it.
• Google says not to expect every URL to be indexed, only canonical pages, so compare the indexed count with your own list of pages you want to rank.
• robots.txt is not a way to keep a page out of Google. A noindex tag is, and Google can only see it on a page it is allowed to crawl.
• The Core Web Vitals pass marks are an LCP of 2.5 seconds, an INP of 200 milliseconds and a CLS of 0.1, measured at the 75th percentile of real visits.
• Every check below has a pass condition. Where the condition is our own rule of thumb rather than Google documentation, the table says “our rule”.
• This page is the self-audit list. The SEO audit template is the spreadsheet you fill in, and the technical SEO audit guide explains what a full audit covers and why.

What does this checklist cover, and how long does it take?

It covers six areas in 42 checks, and a small site takes two to four hours. The first section needs only Search Console and a browser. The later sections go faster with a crawler, and Screaming Frog’s free version crawls up to 500 URLs, which covers most small business sites (Screaming Frog).

Section	Checks	What it answers	Tools you need
1. Indexation	9	Is Google keeping the pages you want to rank?	Search Console, a browser
2. Duplication and canonicals	8	Does each page have one address?	A browser, a crawler on bigger sites
3. Redirects and lost pages	5	Is value leaking through old URLs?	A crawler, your analytics export
4. Structure and internal links	6	Can Google reach every page that matters?	A crawler
5. On-page	7	Does each page say clearly what it is about?	A crawler or page source
6. Speed and page experience	7	Do real visits pass Core Web Vitals?	Search Console, PageSpeed Insights

If you would rather record findings in a sheet, the same thinking is laid out row by row in the free SEO audit template (CSV).

Is Google indexing the pages you want to rank?

Open Search Console, go to Indexing, then Pages, and compare the indexed total with your own count of pages that should rank. Google describes “Discovered - currently not indexed” as a page it found but has not crawled yet, usually because crawling it was expected to overload the site. It describes “Crawled - currently not indexed” as a page it crawled and did not index (Google Search Console Help).

robots.txt controls crawling. Google says it is not a mechanism for keeping a page out of search, and that noindex or a password is (Google Search Central). A noindex only works if the page is not blocked in robots.txt, because Google has to crawl the page to see the tag (Google Search Central).

Google also asks that a sitemap list the URLs you want in search, be submitted in Search Console, and can be declared in robots.txt with a Sitemap line (Google Search Central). A manual action is a human reviewer’s finding that pages break Google’s spam policies, and it shows in the Manual actions report (Google Search Console Help).

#	Check	Passes when	Where to look
1	Indexed pages against expected pages	The indexed total is close to your own list of pages you want to rank, and any gap is explained	Search Console, Indexing, Pages
2	“Discovered - currently not indexed”	No page you want to rank is in this group	Same report, reasons table
3	“Crawled - currently not indexed”	No page you want to rank is in this group	Same report, reasons table
4	Stray noindex	No noindex in the meta robots tag or X-Robots-Tag header of any page you want to rank, or of any template	Crawler filter, or view source and search “noindex”
5	robots.txt	Nothing you want to rank is disallowed, and no noindexed page is also blocked	yoursite.com/robots.txt, read line by line
6	Sitemap contents	Only canonical, indexable URLs that return 200	The sitemap file, checked with a crawler
7	Sitemap declared and submitted	A Sitemap line in robots.txt and a successful submission in Search Console	robots.txt, Search Console Sitemaps report
8	Submitted against indexed	The gap between submitted and indexed sitemap URLs is explained	Page indexing report filtered to the sitemap
9	Manual actions and security issues	Both reports say no issues detected	Search Console, Security and Manual Actions

Staging noindex left in place after a launch is the most expensive single mistake on this list in the accounts we have opened. It removes whole templates at once and nothing on the site looks broken.

Does each page have exactly one address?

It should. Google recommends a rel=“canonical” on the canonical page itself. It calls a redirect a strong signal of which URL should be canonical, and it prefers HTTPS over HTTP versions of the same page (Google Search Central). Google treats URLs as case sensitive, so /Services and /services are two different URLs, and it warns that parameters can multiply URLs pointing at the same content (Google Search Central).

On paginated lists, Google says not to use page one as the canonical for the whole series and to give each page its own canonical instead (Google Search Central). Google also asks for a distinct title on every page (Google Search Central) and a meta description unique to each page (Google Search Central).

#	Check	Passes when	Where to look
10	Self-referencing canonical	Every indexable page has a canonical pointing at itself	View source, or a crawler’s canonical column
11	Protocol and www	http, https, www and non-www all end on one version through a single 301	Type each variant into a browser
12	Trailing slash	With and without the slash resolve to one version	Type both into a browser
13	Letter case	Uppercase variants redirect or canonicalise to the lowercase URL	Type an uppercase version of a key URL
14	Duplicate titles	No two indexable pages share a title	Crawler, page titles report
15	Duplicate meta descriptions	No two indexable pages share a description	Crawler, meta description report
16	Parameter URLs	Filtered and sorted URLs canonicalise to the clean URL	Crawler, URLs containing ”?”
17	Pagination	Each page in a series has its own self-canonical, not one pointing at page one	View source on page two and three

Are old URLs and redirects losing you traffic?

Often they are, after a redesign or migration nobody audited afterwards. Google’s crawlers follow up to 10 redirect hops (Google Search Central), so a chain rarely breaks outright, but each hop is one more thing to maintain. Our rule is one hop. Google’s site move guidance is direct about the homepage: do not redirect many old URLs to one irrelevant destination such as the home page (Google Search Central).

#	Check	Passes when	Where to look
18	Internal links to 404s	Zero internal links return a 404	Crawler, response codes, inlinks
19	Redirect chains	Every redirect reaches its final page in one hop (our rule)	Crawler, redirect chains report
20	Bulk redirects to the homepage	Old URLs redirect to the closest relevant page, not the homepage	Crawl your old URL list
21	Top landing pages from before	Your top 50 URLs from the last analytics export return 200 or redirect once to a relevant page (our rule)	Analytics landing pages export, crawled in list mode
22	Backlinks pointing at 404s	Each is redirected to the closest live page	Search Console Links report, then check each target

Photo: olia danilevich on Pexels.

Can Google reach every page that matters through internal links?

It can if every page you care about is linked from at least one other page, which is Google’s own wording (Google Search Central). The same page says Google can only crawl a link that is an <a> element with an href. It also says anchor text should be descriptive and tell people and Google something about the page it points to. Links added with JavaScript are fine if they follow those rules (Google Search Central).

The numbers in this table for link counts and click depth are our rules of thumb from client work. Google has not published thresholds for either.

#	Check	Passes when	Where to look
23	Orphan pages	No page you want to rank has zero internal links pointing at it	Crawler compared with the sitemap
24	Internal links per important page	Each service or product page has at least three inbound internal links (our rule)	Crawler, inlinks column
25	Click depth	Pages that bring enquiries sit within three clicks of the homepage (our rule)	Crawler, crawl depth column
26	Descriptive anchors	No “click here” or “read more” as the only text of a link	Crawler, anchor text report
27	Varied anchors	Links into a page use natural variations, not one exact phrase everywhere (our rule)	Crawler, inlinks for one page
28	Real links	Navigation and in-content links are <a href> elements, not click handlers	View source or the rendered HTML in URL Inspection

On-page checks: where most audits start, and should not

On-page work matters once the pages are indexed and reachable. Google’s title guidance sets no length limit on a title and says it is truncated in results to fit the device width (Google Search Central). That makes the 60 characters below a practitioner rule for keeping the important words visible. Google’s starter guide says heading order does not matter to Search, though semantic order helps screen readers, and that keyword stuffing is against its spam policies (Google Search Central). For images, the W3C says decorative images should carry an empty alt attribute so screen readers skip them (W3C WAI).

#	Check	Passes when	Where to look
29	One h1	Each page has one h1 that names its subject (our rule, for readers and screen readers)	Crawler, h1 column
30	Heading order	Levels descend without skipping, so an h2 is not followed by an h4 (accessibility, not a ranking rule)	Crawler, or a browser accessibility checker
31	Title tags	Unique, lead with the page’s subject, and keep the key words inside about 60 characters (our rule)	Crawler, page titles report
32	Meta descriptions	Unique, written for the reader rather than copied from the title	Crawler, meta description report
33	Image alt text	Informative images have a description, decorative images have alt=""	Crawler, images missing alt
34	Primary term placement	The term appears in the title, the h1 and the opening copy without repetition for its own sake	Read the page
35	One page per primary term	No two pages target the same main search (our rule)	Your keyword to page map

Do your pages pass Core Web Vitals on real visits?

Check the Core Web Vitals report in Search Console, which is built from real-world usage data and splits mobile from desktop (Google Search Console Help). The good thresholds are an LCP within 2.5 seconds, an INP of 200 milliseconds or less and a CLS of 0.1 or less, measured at the 75th percentile of page loads (web.dev).

Keep this section in proportion. Google says its systems always seek to show the most relevant content even if page experience is sub-par. It also says that chasing a perfect score just for SEO may not be the best use of your time (Google Search Central). Pass the thresholds, then move on.

#	Check	Passes when	Where to look
36	Largest Contentful Paint	2.5 seconds or less on mobile field data	Search Console, Core Web Vitals, Mobile
37	Interaction to Next Paint	200 milliseconds or less	Same report
38	Cumulative Layout Shift	0.1 or less, and you know what moves if it fails	Same report, then PageSpeed Insights
39	Images	Modern format, served near the size they display, with width and height set	PageSpeed Insights, page source
40	Fonts	Self-hosted or preloaded, and none block the first paint	PageSpeed Insights diagnostics
41	Third-party scripts	Every script has a known owner and purpose	Tag manager container and page source
42	Content in the rendered page	The main text and links appear in the rendered HTML Google sees	URL Inspection, view crawled page

Photo: Jakub Zerdzicki on Pexels.

Why does the order matter more than the list?

Because a fix only pays when the page it touches is in the index. Many published SEO checklists open with keyword research, which is the most interesting section and the least useful one while part of the site is excluded.

We have opened accounts where months of new content sat behind a noindex tag carried over from a staging site. Nobody had checked, because checking indexation is dull and writing content is not. Work top down.

If section one is clean, in our experience most of the remaining gains come from sections four and five, internal links and on-page. Speed matters, but once the thresholds pass, extra points buy little. If traffic has already fallen and you are looking for a cause rather than doing a routine check, start with what to check when organic traffic drops instead.

The bottom line

An audit is a list of decisions, not a list of observations. Give every check a pass condition, work from the index outwards, and stop after the first section if it fails, because nothing further down the list can help a page Google is not keeping.

Questions people ask about this