SEO Audit Checklist for Small Sites: Forty Checks, Prioritized

What should a small site check first in an SEO audit?

Check whether your important pages can be found and indexed before you touch titles, descriptions or speed, because nothing else counts if a page is blocked or missing from the index. This page ranks forty checks in five tiers so you fix blockers first and skip what does not apply to your site. The order is Kanso's analysis of two published audit guides, and the method is stated below.

• Tier 0 has 3 setup checks, tier 1 has 10 blockers, tier 2 has 12 checks that make your signals agree, tier 3 has 8 page-level checks and tier 4 has 7 conditional ones.
• Moz's guide says the question of whether a URL actually appears on Google is perhaps the most important item on its list, and that verified indexing lowers the need to dig into the checks after it.
• Moz reports that Google ignores meta descriptions 63% of the time, so description polish sits near the bottom of the order.
• Every row ends with a “skip or defer if” note, which is the part no source supplies.
• The evidence is two audit guides. Where a rule matters to your decision, check it against Google Search Central.

What is an SEO audit?

Serpstat's guide describes the audit as both a check for internal problems and a way to measure how the site performs in search results. Moz frames its version around two questions: can this page rank, and does it meet minimum technical requirements?

For a small site the useful version is narrower than the agency version. Moz says agencies sometimes run audits across hundreds of pages, while a site owner can check a single page by hand in an hour or two and use a crawl to scale beyond that.

How were the forty checks chosen and ranked?

The forty checks come from two audit guides, Moz's technical SEO audit checklist and Serpstat's SEO audit guide, and the ranking is Kanso's own, based on five stated criteria. Both pages were read as extracted text on 20 October 2026.

Ten results appear for the query “seo audit checklist”, and this page used three of them. One was a Notion template with almost no readable text, and the others, including pages from Sitebulb, Yoast, HubSpot and Ahrefs, were not read for this page. A check that appears only in those pages is therefore missing here.

Where both guides cover the same check, such as robots.txt, sitemaps, canonicals and titles, the two were merged into one row. Serpstat's target audience portrait (age, gender, income) was dropped because it is market research, not a check on the site.

1. Blocker. If this fails, can the page still appear in search at all?
2. Dependency. Do later checks only make sense once this one passes?
3. Reach. Does it affect every page, or only some sites?
4. Evidence of effect. Does a source say engines ignore, rewrite or no longer support it?
5. Effort. Can a founder run it without a developer or a paid tool?

A check lands in the earliest tier whose criterion it meets. Blockers and dependencies go first, and conditional checks go last.

Moz's spreadsheet rates each audit point High, Medium, Low or N/A, but the text read for this page does not list those ratings item by item. The tiers here are therefore not Moz's, and the skip notes are not Serpstat's.

What are the five tiers, at a glance?

The forty checks fall into five tiers: three setup checks, ten blockers, twelve signal checks, eight page-level checks and seven conditional ones. Work from the top and fix each tier's failures before starting the next.

The tier 4 checks begin only when the trigger in their “skip or defer if” note applies to you.

Tier 0: Which free accounts do you need before you start?

You need a Google Search Console property, ideally a Bing Webmaster Tools property, and some form of analytics, because the later checks read from them. Moz says these are not part of the audit itself but deliver better data and make the audit easier.

Search Console is the one that cannot be skipped. If you are new to it, the walkthrough of which Search Console reports a founder needs and which to skip covers the reports used in checks 4, 13, 16 and 20, and the Search Console itself is free.

Moz notes analytics is not a ranking factor. Its use in an audit is a baseline: without before-numbers you cannot tell later whether a fix changed anything.

Tier 1: Can search engines reach, read and index your pages?

Search engines can only index pages they can reach and read, so tier 1 checks status codes, directives, robots.txt, internal links and rendering, and it is the tier to finish before anything else. A page that is blocked or missing from the index cannot earn traffic however good its title is. Moz puts the question first on its crawling and indexing list: does the URL appear on Google at all?

Run checks 4 to 9 on your home page and your three most important pages, then run checks 10 to 13 across the whole site. Moz adds that if important content is already indexed, you may not need to spend long on the indexation issues that follow, though they matter more as pages are added at scale.

What is the one robots.txt mistake that can remove a whole site from search?

The costliest mistake is a Disallow: / rule left in the live robots.txt, which tells crawlers to stay away from the entire site. Serpstat says one edit that turns Allow: / into Disallow: / can leave a site unindexed or, worse, dropped from the index with its positions lost.

A typical path to it, in Kanso's reading, is a development copy that was blocked on purpose and a file that was copied to production at launch. Serpstat advises closing test subdomains for developers and QA in robots.txt, which is exactly why the live file needs a read after every deploy.

Suppose a hypothetical 25-page site, invented for illustration and not a case study, fails check 8 because of this one line. No tier 2 or tier 3 fix would matter until it is removed, which is the reason the order exists. If check 13 instead shows pages Google knows about but has not indexed, the guide to finding which indexing stage is broken is the next step.

Tier 2: Do your sitemaps, canonicals and redirects point at the same URLs?

Search engines weigh several signals about which URL is the real one, so tier 2 checks that your sitemaps, canonicals, redirects and links all name the same URL. When the signals conflict, the choice is left to the engine. Moz lists the basic canonical rules: every URL carries a canonical, duplicate-content URLs share one, and alternative URL patterns redirect to it.

Moz also says an XML sitemap is not technically required for ranking but makes discovery easier, and that a page cannot rank if it is never crawled, which is the reason for checks 14 to 16. Placing the sitemap in robots.txt and submitting it directly to each engine are separate steps with separate benefits.

Check 20 deserves care because a canonical tag is a signal, not a command. Moz warns that defining a canonical does not mean Google respects it, since redirects, URL patterns and links also count, and it points to the Search Console status that says Google chose a different canonical than the user.

Check 21 catches a conflict you can create yourself. Moz's rule is to avoid mixed signals, so an indexable page should never have noindexed pages canonicalized to it.

If traffic has already fallen, a full audit is the slower route. The seven-step diagnostic for an organic traffic drop narrows the cause first, and this checklist then covers the part it points to.

Tier 3: Are your titles, descriptions and page speed worth fixing?

Titles, descriptions and speed are worth fixing once tiers 1 and 2 pass, because they change how an indexed page looks and loads, not whether it can appear. Serpstat calls checking titles and descriptions the first step in page analysis and the essential one.

To read them, right-click a page in Chrome and view the source, or use an extension. Serpstat names the SEO META extension and describes its own Website SEO Checker extension as free.

On titles, Serpstat's advice is to describe the page accurately, put the main phrase in once and write for a person. It also says search engines may remove or alter titles stuffed with symbols and promotional wording.

On speed, Serpstat names images as the main cause of slow pages, so fix those first. It also says speed work on larger sites is hard because new page types often need database queries nobody planned for, which is why check 33 samples templates and not every page.

Do meta descriptions deserve audit time?

Only a little: check that key pages have an accurate description, then move on. Moz says the data show Google ignores meta descriptions 63% of the time, and the text read for this page does not state how that figure was measured, so treat it as Moz's figure.

Serpstat gives a second reason for restraint. It says Google snippets run to about 320 characters, that search engines replace descriptions with text found on the page, and that descriptions therefore matter less for optimization. It still says they can raise click-through, and its article is dated October 2023, so the length limit may have changed.

A sensible allocation is to write plain, accurate descriptions for the handful of pages that earn or could earn most of your visits. Leave the long tail alone until the blockers are gone.

Tier 4: Which conditional checks can a small site skip?

Most small sites can skip most of tier 4, because each of its seven checks applies only when a specific situation exists, such as several languages, paginated lists or a past penalty. The table states each trigger.

Checks 38 and 39 are closer to growth work than to a site audit. Competitor tracking is easier as a recurring routine, and the weekly growth checklist sets one out.

Serpstat also says content review cannot be handed entirely to software, because a human auditor still has to work with the texts. That is why this list has no content-quality check: a person has to judge it.

In what order should you run the checks, and when should you stop?

Run them in tier order, fix each tier's failures, and stop when the only checks left are conditional ones whose trigger does not apply to you. The process below is Kanso's, adapted from the Pass, Fail and N/A marking that Moz's spreadsheet uses.

1. Set up tier 0. If Search Console shows no data yet, wait for it before judging checks 4 and 13.
2. Run checks 4 to 9 on the home page and three key pages. Fix any failure before moving on.
3. Run checks 10 to 13 across the whole site.
4. Run tier 2 on one page per template, plus the whole-site checks 14 to 16, 19 and 25.
5. Run tier 3 on key pages only, with check 29 last.
6. Open tier 4 only when a trigger applies.
7. Log each result in a sheet with the check number, Pass, Fail or N/A, a note and the action, as Moz's spreadsheet does.

How long this takes is not documented for a whole site. Moz says one page can be audited by hand in an hour or two, and for a site of hundreds of pages the practical route is to check one page per template and use a crawl for the rest.

Repeat tier 1 after any launch, redesign or migration, since those are the moments when a staging rule or a changed URL can undo earlier work. That timing is Kanso's suggestion, supported by Serpstat's warning about robots.txt edits and Moz's guidance on redirects and canonicals.

What is documented and what is Kanso's analysis?

The checks and figures come from Moz and Serpstat, while the tiers, skip notes and process are Kanso's analysis. The lists keep the two apart.

Documented, as read on 20 October 2026:

• Moz says a single page can be audited by hand in an hour or two, and that a crawl helps when auditing many pages.
• Moz says Google ignores meta descriptions 63% of the time. It does not state a method in the text read.
• Moz says high-priority pages should be within 4 to 5 clicks of the home page, and that a canonical tag is one of several signals Google uses.
• Serpstat says one edit from Allow: / to Disallow: / can drop a site from the index, and that a 503 lasting more than two days is treated as a 404.
• Serpstat says images are the main cause of slow loading, that AMP pages stopped being served via the Google domain in 2023, and that Google snippets run to about 320 characters. Its article is dated October 2023.

Kanso's analysis:

• The five tiers and the order of the checks within them.
• Every “skip or defer if” note.
• The stop rule, the process order and the advice to repeat tier 1 after launches and migrations.
• The hypothetical 25-page site, which is invented to show the order and reports no result.
• The selection of checks for launches, redesigns and migrations, which neither source published as a list.

Where does the Kanso SEO Agent fit into an SEO audit?

The SEO Agent is the part of Kanso that does audit-style work: it audits your site and Search Console data and turns what it finds into recommendations and tracked tasks for your approval. Read the SEO Agent page for what it covers, and pricing for plans.

This checklist is a manual process and is not a description of what the agent checks. Kanso does not promise rankings, so a clean audit means the basics are in place, not that traffic follows.

SEO audit checklist: frequently asked questions

Sources

• The Technical SEO (& Beyond) Site Audit Checklist, Moz
• Advanced SEO Audit: A Complete Guide to All Stages of The Analysis, Serpstat Blog (dated 23 October 2023, with a July 2024 update note)