Google publishes three minimum technical requirements. A typical technical SEO audit reports sixty findings. The gap between those numbers is where most of the money in this field is spent.

The three are worth knowing exactly, because everything beyond them is optimisation rather than eligibility, and because the documentation is unusually candid about which of its own guides most sites should skip.

This page sets out the requirements, the size thresholds at which the advanced advice starts applying, and the one configuration mistake that keeps pages in the index against their owner’s wishes.

The three requirements, verbatim

They appear together on one page, with an explicit caveat attached.

Requirement one. Googlebot is not blocked. The documentation adds that Google only indexes pages that are accessible to the public and which do not block its crawler from crawling them.

Requirement two. The page works, meaning that Google receives an HTTP 200 success status code.

Requirement three. The page has indexable content, defined as textual content in a file type Google Search supports, that does not violate the spam policies.

The caveat, verbatim. Just because a page meets these requirements does not mean that a page will be indexed; indexing isn’t guaranteed.

And discoverability, from the companion page. Some pages are known because Google has already visited them, others are discovered when Google extracts a link from a known page, and still others when you submit a sitemap.

What that gives you. A four-item checklist: reachable by a link or a sitemap, not blocked, returns 200, has real content. Everything in a sixty-point audit that is not one of those four is a refinement, and should be justified as such.

The guide Google tells you not to read

The crawl budget documentation is the clearest example of the platform being more relaxed than the industry built on it.

The opening, verbatim. That the guide describes how to optimise crawling of very large and frequently updated sites, and that if your site does not have a large number of pages that change rapidly, or if your pages seem to be crawled the same day that they are published, you do not need to read this guide.

What it says to do instead. For Google Search specifically, keeping your sitemap up to date and checking the page indexing report regularly is adequate.

The thresholds, verbatim. Large sites of 1 million or more unique pages with content that changes moderately often, about once a week. Medium or larger sites of 10,000 or more unique pages with very rapidly changing content, daily. Plus sites with a large portion of URLs classified as discovered but not currently indexed.

The disclaimer attached to those numbers. That they are a rough estimate to help you classify your site, and these are not exact thresholds.

Why this matters commercially. Crawl budget optimisation is a standard line item in technical SEO proposals. For a B2B site with a few hundred pages, the platform’s own documentation says it does not apply.

The honest exception. The third condition. If Search Console shows a large share of your URLs as discovered and not indexed, you have a real problem, regardless of your page count.

The three published minimum technical requirements and the thresholds governing advanced crawl guidanceSummary of the three minimum technical requirements published by the search platform for a page to be eligible for indexing, together with the site size thresholds at which its advanced crawl budget guidance becomes applicable. The three requirements are that the platform’s crawler is not blocked, meaning the page is accessible to the public and does not block the crawler from crawling it; that the page works, meaning the platform receives an HTTP two hundred success status code; and that the page has indexable content, meaning textual content in a supported file type that does not violate the spam policies. The documentation attaches an explicit caveat stating that meeting these requirements does not mean a page will be indexed, since indexing is not guaranteed. A companion page adds the discovery dimension, noting that pages become known either because the platform has already visited them, or because it extracts a link to them from a known page, or because the operator submits a sitemap. Regarding advanced guidance, the crawl budget guide opens by stating that it describes optimisation for very large and frequently updated sites and that operators whose sites lack a large number of rapidly changing pages, or whose pages appear to be crawled the same day they are published, do not need to read it, adding that keeping the sitemap current and checking the page indexing report regularly is adequate for such sites. The thresholds it publishes are large sites of one million or more unique pages with content changing about weekly, medium or larger sites of ten thousand or more unique pages with content changing daily, and sites with a large portion of their addresses classified as discovered but not currently indexed. The documentation describes these figures as a rough estimate to help classify a site and states explicitly that they are not exact thresholds.Three requirements, and one caveat1. Not blockedPublicly accessible, crawlernot disallowed.2. Returns HTTP 200”The page works.”3. Indexable contentSupported file type, notviolating spam policies.”Just because a page meets these requirements doesn’t mean that a page will be indexed.”And the guide that opens by telling you not to read it”If your site doesn’t have a large number of pages that change rapidly, or if your pages seem to becrawled the same day that they are published, you don’t need to read this guide.”When crawl budget does applyLarge sites, content changing weekly1,000,000+ unique pagesMedium or larger, content changing daily10,000+ unique pagesOr: many URLs “discovered, not indexed”any size, and this one is real
The crawl budget guide opens by telling most sites not to read it, and publishes the page counts where it applies. Source : Google Search Central technical requirements and crawl budget guide (2026)

The mistake that keeps pages in the index

One configuration error accounts for a large share of “why is this still in Google” tickets, and the documentation explains it precisely.

What robots.txt does, verbatim. It is used mainly to avoid overloading your site with requests, and it is not a mechanism for keeping a web page out of Google.

What happens if you use it that way. A page disallowed in robots.txt can still have its URL appear in Google Search results if other pages link to it with descriptive text, because Google can index the URL without visiting the page.

The clause that explains the whole failure, verbatim. That robots meta tags and X-Robots-Tag headers are discovered when a URL is crawled, so if a page is disallowed from crawling in robots.txt, then any information about indexing or serving rules will not be found and will therefore be ignored.

Stated as a rule. If an indexing rule must be followed, the URL containing that rule cannot be disallowed from crawling.

So the common sequence is self-defeating. Add noindex to a page, then block it in robots.txt to be doubly safe. The block prevents Google from ever seeing the noindex, and the page stays.

What to do instead, in the documentation’s own words. To keep a page out of Google, block indexing with noindex and allow Google to crawl the URL. Or password-protect it, or remove it entirely.

The corollary for audits. A finding of “page not blocked in robots.txt” is not automatically a problem. On any page carrying a noindex, it is a requirement.

Sitemaps, with the number attached

Another item that appears in every audit and has a published threshold.

The general position, verbatim. If your site’s pages are properly linked, Google can usually discover most of your site. A sitemap helps search engines discover URLs but does not guarantee that all items in it will be crawled and indexed. In most cases a site will benefit from having one.

When you might not need one, verbatim. If your site is small, by which they mean about 500 pages or fewer, counting only pages you think need to be in search results.

When you likely do. A large site, a new site with few external links, or a site with a lot of rich media or appearing in news.

The technical limits. 50 MB uncompressed, or 50,000 URLs, per sitemap file.

How to read that in a B2B context. Most B2B sites are well under 500 pages. A sitemap is still worth having because it is free and it feeds the indexing report, but its absence is not the cause of your traffic problem.

And what it never does. Force indexing. The documentation is explicit that inclusion in a sitemap guarantees nothing.

Why blocking a page in the robots file prevents its no-index instruction from being appliedExplanation of why blocking a page in the robots file prevents the no-index instruction on that page from ever being applied, which is the most common self-defeating configuration in technical search optimisation. The documentation states that the robots file is used mainly to avoid overloading a site with requests and is not a mechanism for keeping a web page out of the search index. It further states that a page disallowed in the robots file can still have its address appear in search results if other pages link to it with descriptive text, because the search engine can index the address without visiting the page. The decisive clause is that robots meta tags and X-Robots-Tag headers are discovered only when an address is crawled, so if a page is disallowed from crawling through the robots file then any information about indexing or serving rules will not be found and will therefore be ignored, stated as a rule that if indexing or serving rules must be followed then the addresses containing those rules cannot be disallowed from crawling. The consequence is that the common sequence of adding a no-index instruction to a page and then additionally blocking that page in the robots file for extra safety is self-defeating, since the block prevents the search engine from ever reading the no-index instruction and the page therefore remains in the index. The documented remedy is to block indexing with the no-index instruction while allowing the search engine to crawl the address, or alternatively to password-protect the page or remove it entirely. A corollary for audit reports is that a finding of a page not being blocked in the robots file is not automatically a defect, and on any page carrying a no-index instruction it is in fact a requirement.Why your noindex is being ignoredThe self-defeating sequence1. Add noindex to the pageCorrect so far.→2. Also block it in robots.txt”To be doubly safe.”→3. The page stays indexedBecause step 2 hid step 1.The documentation’s own explanation”If a page is disallowed from crawling through the robots.txt file, then any information aboutindexing or serving rules will not be found and will therefore be ignored.”And robots.txt alone does not remove itA blocked URL can still appear if other pages link to it.What to do insteadnoindex, and allow crawling. Or password-protect it.Audit corollary: “not blocked in robots.txt” is a requirement on any page carrying a noindex, not a defect.
The most common self-defeating configuration in technical SEO, explained by the documentation itself. Source : Google robots.txt and robots meta tag documentation (2026)
Published guidance on when a sitemap is needed and its technical limitsSummary of the search platform’s published guidance on when a sitemap file is necessary, together with its technical limits. The general position states that if a site’s pages are properly linked the platform can usually discover most of the site, that a sitemap helps search engines discover addresses but does not guarantee that all items within it will be crawled and indexed, and that in most cases a site will nonetheless benefit from having one. The documentation states that an operator might not need a sitemap if the site is small, defining small as approximately five hundred pages or fewer, and specifying that only pages the operator believes need to appear in search results count toward that total. It states that an operator might need one if the site is large, if the site is new and has few external links pointing to it, or if the site contains a large quantity of rich media content such as video and images or appears in the platform’s news product. The technical limits published for a single sitemap file are fifty megabytes uncompressed or fifty thousand addresses, whichever is reached first. The practical reading for a business-to-business site is that most such sites fall well below five hundred pages intended for search results, so a sitemap remains worth having because it costs nothing and feeds the indexing report, but its absence is unlikely to be the cause of a traffic problem, and its presence never forces indexing since the documentation states explicitly that inclusion guarantees nothing.Sitemaps, with the published numberYou might not need one”Your site is ‘small’. By small, we meanabout 500 pages or fewer”Counting only pages you think need to bein search results.You likely doA large site.A new site with few external links.A lot of rich media, or news.Technical limits per file50 MB uncompressed, or 50,000 URLs.What it never doesForce indexing. Inclusion guarantees nothing.Read for a B2B siteMost are well under 500 pages. Have one anyway, because it is free and it feeds the indexing report.But its absence is not why your traffic is flat.
About 500 pages, counting only what you want in results. And it guarantees nothing either way. Source : Google sitemap documentation (2026)

What the documentation says does nothing

Short list, and the boundary of it is worth knowing.

Meta keywords, verbatim. That the meta keywords tag is not used by Google Search, and it has no effect on indexing and ranking at all. That is documented, quotable and unambiguous.

Keyword density. Not addressed anywhere in the search documentation. The widely repeated statement that it is not a factor comes from spokesperson remarks in public forums, not from a page you can cite.

Third-party authority scores. Same situation. The well-known statement that Google does not use them is a spokesperson position expressed publicly, and it is not in the documentation.

Why that distinction matters. In a client argument, “Google’s documentation states it has no effect at all” and “a Google employee said so on a forum” are different weights of evidence. Only one of the three items above carries the first.

What that implies for audits generally. Ask which claims in the report cite documentation, which cite a spokesperson, and which cite nothing. That single question usually reduces sixty findings to a workable list.

And one thing worth saying plainly. None of this means technical work is unnecessary. It means the requirements are few, the thresholds are published, and most of what fills an audit is optimisation that should be justified against the traffic it can plausibly move.

Three tiers of evidential support behind typical technical search optimisation findingsFramework separating the findings in a typical technical search optimisation report into three tiers according to the evidence supporting each. The first tier comprises claims documented directly by the search platform, which can be quoted with an address and defended in a client discussion; examples include the statement that the meta keywords tag is not used by the platform’s search and has no effect on indexing and ranking at all, the three minimum technical requirements for indexing eligibility, the crawl budget size thresholds, the approximately five hundred page guidance for sitemap necessity, and the explanation that indexing rules on a page cannot be read if that page is disallowed from crawling. The second tier comprises positions expressed publicly by platform spokespeople in forums, social media or recorded discussions but not present in the documentation; the widely repeated statements that keyword density is not a ranking factor and that the platform does not use third-party domain authority scores both fall into this tier, and neither appears anywhere in the search documentation. The third tier comprises findings that cite nothing at all, which constitute the bulk of a sixty-point audit and represent optimisation conventions rather than documented requirements. The practical instruction is to ask, of each finding in a report, which tier supports it, since that single question typically reduces a sixty-item list to a workable set. This does not imply that technical work is unnecessary, only that the genuine requirements are few, the applicable thresholds are published, and the remainder should be justified against the traffic it can plausibly influence.Three tiers of evidence in one reportTier 1: documented, quotable with a URL”The meta-keyword tag is not used by Google Search, and it has no effect on indexing and ranking at all.”The three requirements. The crawl budget thresholds. The ~500-page sitemap line. The robots/noindex conflict.Tier 2: a spokesperson said it publicly, but it is not documentedKeyword density is not a factor. Third-party authority scores are not used.Probably true. Not citable as documentation, and worth less in an argument.Tier 3: cites nothingThe bulk of a sixty-point audit. Optimisation convention rather than documented requirement.Not necessarily wrong. Just needs justifying against the traffic it can plausibly move.One question per finding: which tier is this? Sixty items becomes a list you can actually work through.
Ask which findings cite documentation, which cite a spokesperson, and which cite nothing. Sixty becomes a workable list. Source : Google Search Central documentation (2026)

What to do with an audit you have been handed

Six steps, and the first two dispose of most of it.

Check the four eligibility items first. Reachable, not blocked, returns 200, has content. If any page you care about fails one of those, that is the whole priority list until it is fixed.

Sort every remaining finding by tier. Documented, spokesperson, or nothing. Then work only the first group until it is empty.

Check the discovered-but-not-indexed report. This is the one crawl condition that applies at any size, and it is the only crawl-related finding on a small site that is likely to be real.

Verify that no page you want removed is blocked from crawling. The single most common cause of a page refusing to leave the index.

Ignore crawl budget unless you cross the published thresholds. A million pages changing weekly, or ten thousand changing daily. If you have four hundred pages, this is not your problem and paying for it is a transfer, not an investment.

Then ask what the remaining items are expected to move. Not whether they are best practice. What traffic, on which pages, by when. Anything that cannot answer goes to the bottom.

A triage procedure for prioritising the findings in a technical search optimisation auditProcedure for triaging the findings contained in a technical search optimisation audit report, in six ordered steps of which the first two dispose of most of the content. The first step is to check the four eligibility items, namely that each page is reachable by a link or sitemap, that the crawler is not blocked from it, that it returns an HTTP two hundred status code and that it contains indexable content; where any page of importance fails one of these, that constitutes the entire priority list until resolved. The second step is to sort every remaining finding by the tier of evidence supporting it, whether documented by the platform, stated publicly by a spokesperson, or citing nothing, and to work only the documented group until it is exhausted. The third step is to examine the discovered but not currently indexed report, since this is the sole crawl-related condition that applies at any site size and is the only crawl finding on a small site likely to represent a genuine problem. The fourth step is to verify that no page intended for removal is blocked from crawling, since that configuration is the single most common cause of a page refusing to leave the index. The fifth step is to disregard crawl budget entirely unless the site crosses the published thresholds of one million or more pages changing weekly or ten thousand or more changing daily, since paying for crawl budget optimisation on a four hundred page site constitutes a transfer rather than an investment. The sixth step is to ask of each remaining item what traffic it is expected to move, on which pages, and by when, rather than whether it constitutes best practice, with anything unable to answer that question moved to the bottom of the list.Triage, in six steps1. The four eligibility items, first and aloneReachable. Not blocked. Returns 200. Has content. Anything failing these is the whole priority list.2. Sort the rest by tierDocumented, spokesperson, or nothing.3. Discovered, not indexedThe one crawl issue that is real at any size.4. Nothing you want removed is blockedThe commonest cause of a page that will not leave.5. Skip crawl budgetUnless 1M pages weekly, or 10K daily.6. Then ask what each remaining item will moveWhat traffic, on which pages, by when. Not whether it is best practice.On a 400-page site, buying crawl budget optimisation is a transfer, not an investment.
Four eligibility items first. Then sort by what each finding cites. Crawl budget only past the published thresholds. Source : Method, applied to Google's own documentation (2026)

Where to go next

You are moving the site. Website migration without losing rankings.

You are working on speed. Website speed: what matters.

You are choosing a host. B2B website hosting and performance.

You want to know what AI summaries do to clicks. AI Overviews and organic traffic.

You are weighing paid against organic. Paid advertising vs organic.

You are rebuilding. Signs it is time for a B2B website redesign.

In short

  • Three published requirements: Googlebot not blocked, HTTP 200, indexable content. Plus discoverability by link or sitemap.
  • And an explicit caveat: meeting them does not guarantee indexing.
  • The crawl budget guide tells most sites not to read it, and says keeping the sitemap current and checking the indexing report is adequate.
  • Its thresholds are 1,000,000+ pages changing weekly or 10,000+ changing daily, described as a rough estimate rather than exact.
  • A sitemap may be unnecessary below about 500 pages that you want in search results. Limits are 50 MB or 50,000 URLs per file.
  • robots.txt does not remove pages. A blocked URL can still appear if other pages link to it.
  • And blocking hides your noindex. Rules on a page cannot be found if the page cannot be crawled.
  • Only meta keywords is documented as having no effect. Keyword density and authority scores are spokesperson positions, not documentation.

Fix the four eligibility items, then sort the rest by what it cites. Book a diagnostic, or see how we approach B2B websites.