AdSense readiness

Our server fetches your home page, then robots.txt, ads.txt and any sitemap, then up to eight pages in total — the ones your footer links to as policies, and a spread of content pages taken from your sitemap and your own links. It reads what came back and runs every check against it. The whole thing stops after 25 seconds whatever it has.

It obeys your robots.txt for every URL except the one you submitted, which is a page you asked us to fetch. Anything skipped is reported rather than quietly dropped. The report is cached for fifteen minutes, so two people checking the same site cost it one crawl and not two.

The address you type is sent to our API and to the site you name. Nothing else about you is recorded — no account, no cookie set by this tool, no stored history of what was checked.

A checker that calls eleven things blockers has no blockers. So the findings are split by what a failure actually costs, and the top tier is kept small on purpose — 7 of the44 checks are in it.

• Blocker. Something that stops an application on its own: Google cannot reach the site, cannot crawl it, a page Google requires by name is missing, or a policy names the thing outright. These are not judgement calls and there is no version of them that passes.
• Judgement call. Something a reviewer weighs. Strongly correlated with the rejection almost everyone gets — low value content — but a human decision, and a site can pass with several of these open.
• Worth knowing. Real, cheap to fix, and it would never decide a review.

Generated from the same catalogue the checker runs, so this list cannot drift from what it actually does.

Thirty posts. A thousand words each. A domain six months old. A hundred visitors a day. None of those appear in any Google document, and a tool repeating them as requirements is repeating a rumour with a progress bar attached.

What Google does publish is short: applicants must be 18 or over, must be able to access the site's HTML source, and the content must be "high-quality, original, and attract an audience". Its site-readiness guidance adds original content, an accessible navigation bar and a coherent layout. That is close to the whole documented bar.

The rejection almost everyone gets is three words — low value content — and it is a judgement made against experience, expertise, authoritativeness and trust. Word count rarely decides it. A short piece by someone who plainly knows the subject passes where a long generic one fails.

This tool does use thresholds of its own, because "it depends" helps nobody. Every one of them is marked Our judgement in the report, and that badge is drawn differently from the others so it cannot be mistaken for a rule.

This is not a disclaimer. It is the part that decides most reviews, and every checker that stays quiet about it teaches its readers that the measurable part is the whole part.

• Whether the writing is original, or reworded from somewhere else. Comparing a site against the rest of the web needs a corpus this tool does not have.
• Whether the person writing knows the subject. Expertise is the thing a reviewer is actually assessing, and it is not a property of the HTML.
• Whether a reader finishes a page better off than they started. That is what "high-quality, original, and attract an audience" means, and it is read, not measured.
• Whether the publishing pace looks like a person or a pipeline. Thirty pages posted in one day is the fingerprint of scaled content abuse whatever each page says.
• Whether images, quotes and data on the site are licensed for use.
• Whether the applicant is over 18 and owns the site — both are stated eligibility requirements and neither is visible from outside.
• What real visitors experience. Core Web Vitals are measured at the 75th percentile of actual page loads over 28 days; one fetch of the HTML from one place cannot approximate it, and the two things that will break yours — the consent platform and the ad slots — do not exist on the site yet.
• Where the traffic comes from. Google’s policies prohibit inventory built for arbitrage or bought traffic, and a report on the pages cannot see who arrives at them.
• Whether any of it is true. Every check here reads how a page is built, not whether what it says is correct — and on a money subject, correctness is most of what a rater is judging.
• Whether the applicant’s own country is eligible. AdSense availability, payment methods and tax requirements follow where the account holder lives, not where the site is hosted or who reads it, and Google has historically required six months’ site ownership before applying in some locations including China and India. None of that is a property of the site.
• What a reviewer will think of the layout. "An inviting layout" is Google’s phrase, and it is a judgement.

The research behind this tool is written up in full, with the sources and the parts that could not be verified marked as such:AdSense requirements: the checklist is folklore, the three-word rejection is real. It covers the whole documented bar, the three words most rejections actually use, the obligations that come from European law rather than from Google, and theads.txt mistake that a single redirect causes.

1. Make sure Mediapartners-Google is not blocked. That is the AdSense crawler, and it is separate from Googlebot. A site that ranks perfectly well can be shut out of ad serving by one line in robots.txt, and nothing about the site looks wrong when it is.
2. Put a real person on the about page. Not a positioning sentence — what they know, from where, and why they are worth reading on this subject. On a money subject, an anonymous operator is close to an automatic rejection, and this is the single highest-value page on the site for approval purposes.
3. Publish a privacy policy that names Google. It is the only page whose contents Google mandates: third-party vendors including Google use cookies to serve ads, and visitors must be pointed at the Ads Settings and aboutads.info opt-outs.