SEO Audit Checklist: The Checks That Matter, With a Pass Line for Each

Here is what nobody selling an SEO audit checklist tells you. A site can tick 48 boxes out of 50 and still lose most of its leads, because the two boxes it failed were the only ones that mattered. Most checklists give a missing image alt attribute the same weight as a noindex tag on your best service page.

I am Armen, an SEO consultant in Barcelona, and I run the agency ACERO Digital with a team of specialists. When a business owner sends me the checklist they downloaded, the problem is rarely that it is wrong. The problem is that every line says “check your canonicals” and not one says what a pass looks like.

So this version is built differently. Every check comes with a pass line and the report where you verify it. At the end there is a section on the popular items I would cross off your list entirely.

What should an SEO audit checklist include?

An SEO audit checklist should include only checks with a measurable pass line: a threshold or a plain yes or no that two different people would agree on. If a line says “optimise your content”, it is advice, not a check.

Before you start, write down the ten to twenty URLs that bring in money or enquiries. A failure on those pages is urgent. The same failure on a tag archive from 2019 can wait, or be ignored forever.

You need very little to run it. Google Search Console, PageSpeed Insights, a crawler such as Screaming Frog (free up to 500 URLs), and a spreadsheet with two columns: pass or fail.

Can Google crawl and index your important pages?

This is the first block because a page that is not indexed cannot rank, whatever else you fix. Pass every check here before you spend an hour anywhere else.

1. Money pages are indexed. Pass: URL Inspection in Search Console says “URL is on Google” for every page on your list. The site: operator is not a reliable substitute.
2. No stray noindex. Pass: none of your commercial URLs appear under “Excluded by noindex tag” in the Page indexing report. Staging settings that survive a launch are the classic cause.
3. Robots.txt blocks only what you meant. Pass: open yourdomain.com/robots.txt in a browser and every Disallow line has a reason you can explain. Remember that blocking a URL in robots.txt does not remove it from the index; it only stops Google reading the page, including any noindex tag on it.
4. Commercial pages are not stuck in limbo. Pass: none of them sit under “Crawled, currently not indexed” or “Discovered, currently not indexed”. The first usually means Google read the page and found it not worth keeping. The second often points at weak internal linking.
5. The sitemap is clean. Pass: every URL in the XML sitemap returns a 200 status, is indexable and is its own canonical. A sitemap full of redirects teaches Google to trust it less.

Does every page send Google one clear signal?

Each URL should say one thing about itself, consistently, in its canonical tag, its title, its internal links and its redirects. Mixed signals are where most “why is Google showing the wrong page” problems begin.

6. Canonicals agree with you. Pass: in the Page indexing report, “Duplicate, Google chose different canonical than user” is empty for your money pages. If Google overrules you, your internal links or sitemap usually disagree with your tag.
7. One version of the site. Pass: the http and www variants all redirect in a single hop to the one version you use.
8. No redirect chains or broken internal links. Pass: your crawl shows zero internal links pointing to 404 pages and no redirect longer than one hop. Google will follow a chain; your crawl budget and your users pay for it.
9. Titles are unique and describe the page. Pass: no title is shared across several URLs, and each one contains the phrase a buyer would actually search. Forty identical titles is one template bug, not forty tasks.
10. Money pages are easy to reach. Pass: each commercial page sits within three clicks of the homepage and receives links from relevant pages, not only from the menu. That three click line is my rule of thumb, not a Google rule, but orphaned service pages almost never perform.

Is the site fast and usable on a real phone?

The pass line here is Core Web Vitals field data, not a Lighthouse score. Field data comes from real Chrome users over the last 28 days, and a page passes a metric only when 75 percent of visits meet the “good” threshold.

Check	Pass line	Where to look
Largest Contentful Paint	2.5 seconds or less	PageSpeed Insights, field data section
Interaction to Next Paint	200 milliseconds or less	Search Console, Core Web Vitals report
Cumulative Layout Shift	0.1 or less	Search Console, Core Web Vitals report
HTTPS everywhere	No mixed content warnings, valid certificate	Browser padlock, crawler security tab
Mobile content parity	Same text and structured data as desktop	Compare both crawls side by side

11. Test templates, not the homepage. Pass: one representative URL per template, such as your service page and your blog post, meets all three thresholds. The homepage is usually the fastest page on the site and the one your buyer lands on least.
12. Mobile shows the full page. Pass: nothing important is hidden or cut on the phone version. Google indexes the mobile version of your site, so content that only exists on desktop may as well not exist.

A small site with little traffic may show “not enough data” in the field report. In that case use the lab numbers as a rough guide and do not lose sleep over a score in the 80s.

Does your content match what people are actually searching for?

A page passes this block when it looks like the kind of result Google already rewards for its query. No tool can grade this for you, which is exactly why it is the block most checklists rush.

13. Intent match. Pass: search your main query in a private window. If the top results are comparison guides and your page is a pure sales page, that is a fail, however well written it is.
14. No cannibalisation. Pass: in the Search Console Performance report, filter by your main query and check the Pages tab. One URL should collect the clicks. Two of your own pages trading places week to week is a fail.
15. No thin or duplicated pages. Pass: every indexable page has content that exists nowhere else on the site. Location pages that swap “Gràcia” for “Sants” and change nothing else are the usual offender.
16. Content is current. Pass: pages that lost the most clicks over the last six months have been reviewed against what now ranks. Prices and years in titles age faster than anyone expects.

Is your structured data valid and truthful?

Structured data passes when it validates and when every property describes something a visitor can actually see on the page. Markup that claims reviews or prices you do not show is a manual action waiting to happen.

17. It validates. Pass: the Rich Results Test shows no errors on one page per template, and the Schema Markup Validator agrees.
18. The business is described once, clearly. Pass: Organization or LocalBusiness markup carries the same name, address, phone and logo as your site footer and your Google Business Profile.

Do not fail yourself for missing FAQ rich results. Since 2023 Google has shown them only for a small group of well known government and health sites, so their absence on a business site is expected, not a bug.

What does a local business in Barcelona need to add?

A local business needs a second audit layer, because the map results and your Google Business Profile decide many of the calls you get. This is also where a typical downloaded checklist is silent.

19. Primary category is precise. Pass: your Google Business Profile primary category is the most specific one that fits (“Dental clinic”, not “Health”), and the secondary categories cover real services only.
20. Name, address and phone match everywhere. Pass: identical details on your site, your profile, Apple Maps, Bing Places and the main directories you appear on. A carrer written three different ways is enough to blur the signal.
21. Reviews are alive. Pass: new reviews arrive regularly and every recent one has a reply from the owner.
22. Languages are connected. Pass: if you publish in Catalan, Spanish and English, each version declares the others with hreflang, including a self reference, and every tag is returned by the page it points to. One missing return tag breaks the pair.

If this block produces more fails than passes, that is normal, and it is often the fastest win of the whole audit. It is also the part of our local SEO work in Barcelona that clients notice first, because the phone starts ringing before the organic rankings move.

Can AI search engines find and quote your pages?

Your site passes this block when AI crawlers are allowed in and your pages answer questions in sentences a model can lift. More buyers now ask ChatGPT or Perplexity for a recommendation before they search Google, and a checklist that stops at Google misses them.

23. AI search crawlers are allowed. Pass: robots.txt does not block OAI-SearchBot, PerplexityBot or ClaudeBot, unless blocking them is a decision you made on purpose. Security plugins and CDN bot settings sometimes block them for you.
24. You exist in Bing. Pass: your site is verified in Bing Webmaster Tools and your money pages are indexed there. Microsoft Copilot answers from Bing, and it is a free hour of work.
25. Sections answer their own heading. Pass: read the first two sentences under each H2 on your key pages. If they answer the heading without the rest of the page, a model can quote them.
26. Your business is described the same way off site. Pass: your profile, directories and any press mentions describe what you do in consistent words. Models recommend businesses they can describe confidently.

Which SEO audit checklist items can you safely skip?

You can skip any check that has no effect on how Google crawls, indexes, ranks or displays your pages. Plenty of popular list items fall into that bucket, and every hour spent on them is an hour taken from a real fail.

Cross these off:

• Meta keywords. Google has ignored the tag for over fifteen years.
• Keyword density. There is no target percentage. Write for the query and move on.
• Exact title and meta description lengths. There is no official character limit, and Google rewrites many descriptions anyway. Make them clear; stop counting.
• Text to HTML ratio and W3C validation errors. Tools love flagging them. Neither is something Google ranks on.
• Third party authority scores. Domain Rating and Authority Score are useful for comparing sites, but they are not Google metrics and not a pass line.
• A routine disavow file. Google’s systems ignore most spammy links on their own. Reserve disavowing for a manual action or a paid link scheme you know about.
• llms.txt. No major AI search engine has confirmed it uses the file. Checks 23 to 26 do the real work.

This is the part I would defend hardest. A shorter checklist that you actually finish beats a long one that stalls at item 31.

What I would check first

If you only have one hour, run checks 1 to 5 on your list of money pages. They are the ones that can silently zero out a page, and I still find at least one of them failing on sites that look perfectly healthy from the outside.

If you have a day, add the local block and the AI search block, because that is where most small businesses in Barcelona have the biggest gap and the least competition. Then take your fails, sort them by which money pages they touch, and hand the top few to whoever can ship them this month.

And if you would rather have someone run the full SEO audit and turn the fails into a sequenced fix list, that is exactly the work we do. Send me your site and tell me what worries you most about it.