Is it true that Search Console only analyzes a portion of your site’s pages?

Which Search Console reports use sampling?

Mueller specifies that aggregated reports — mobile usability, structured data, Core Web Vitals — do not scan the entirety of your index. Google selects a representative subset of pages to establish these diagnostics.

This sampling approach has been around for a long time, but Google rarely communicates it explicitly. SEO practitioners often focus on the index coverage report, which reflects all known URLs well. The technical quality reports, however, operate differently.

Why does Google do it this way?

From an infrastructure perspective, scanning millions of pages for each type of technical report would be resource-intensive. Sampling allows for quick and updated diagnostics without monopolizing the computing power needed for indexing itself.

For a site with 50,000 pages, Google may analyze only 5,000 in the Mobile Usability report. This ratio is not fixed and varies depending on the size of the site, its update frequency, and probably its notoriety. A small site may be scanned thoroughly, while a large site will undergo more pronounced sampling.

Is a decrease in the number of analyzed pages concerning?

No, and that’s the crux of Mueller's message. If you notice that the Mobile Usability report drops from 8,000 to 6,000 analyzed pages, it doesn’t necessarily mean that 2,000 pages have disappeared from the index or have issues.

The sample may naturally decrease with no correlation to the health of the site. Google adjusts the sample size based on internal criteria — server load, crawl priority, observed site stability. If your site hasn't changed and traffic remains stable, a drop in the number of pages in these reports likely relates to the sampling mechanics, not a warning signal.

• Aggregated reports (mobile, speed, structured data) use a sample, not the complete index
• The sample size fluctuates without systematically reflecting a technical problem
• The index coverage report, however, remains comprehensive and reflects all discovered URLs
• A decrease in the volume of analyzed pages does not necessarily imply deindexation or degradation

Is this statement consistent with ground observations?

Yes, largely. SEO practitioners have long noticed unexplained variations in the technical reports from Search Console. A stable site can show 12,000 pages in the Core Web Vitals report one month and then 9,000 the next month, without structural changes or traffic loss.

What’s lacking is transparency about the sampling criteria. Google does not specify the ratio applied or the criteria for selecting pages. Will a site with 100,000 URLs always be sampled at 10%, 20%, or according to an adaptive logic? [To be verified] — no public data allows for a definitive answer.

What limitations does this approach impose on SEO diagnostics?

The main issue is that a sample can mask localized errors. If 500 pages in a specific section of your site have a structured data problem, but they are not part of the analyzed sample, the report won’t alert you.

Practically, an SEO practitioner cannot solely rely on aggregated reports for a comprehensive audit. It is essential to cross-reference with third-party tools (Screaming Frog, OnCrawl, Botify) that scan the entire site. Search Console becomes a trend indicator, not a complete diagnostic.

In what cases does this rule not apply?

The index coverage report remains exhaustive — all URLs discovered by Googlebot appear there, whether they are indexed, excluded, or in error. It is the source of truth for understanding how many pages Google actually knows about.

Similarly, performance data (impressions, clicks, CTR) in the performance report accurately reflect the site’s complete activity, not a sample. If you see 10,000 clicks, those are 10,000 real clicks, not an extrapolation. Sampling only concerns technical quality reports.

What should you do to compensate for this limitation?

Never rely solely on Search Console for a complete technical audit. Invest in a crawler like Screaming Frog (paid version for larger sites) or a platform like OnCrawl, Botify, or Sitebulb. These tools scan your entire site and detect errors that Google's sampling may miss.

Establish a regular monitoring of key pages. Identify your high-value categories (bestselling product pages, SEA landing pages, pillar content) and manually check their mobile compliance, structured data, and Core Web Vitals. Don't let a random sample dictate what gets checked.

What mistakes should you avoid in interpreting the reports?

Don’t panic if the number of pages in the Mobile Usability report drops by 20% without any other negative signals. First, check organic traffic, the index coverage report, and your rankings on strategic queries. If everything is stable, sampling is likely the cause.

Conversely, don’t rely on the absence of errors in these reports to conclude that everything is fine. A sample may overlook critical issues on entire segments of the site. A complete audit via a crawler remains essential at least once a quarter.

How can I check if my site is being scanned correctly?

Compare the number of pages in the index coverage report (under "Indexed Pages") with the volume of pages in the technical reports. If you have 50,000 indexed pages but only 5,000 in the Core Web Vitals report, you are clearly using a reduced sample.

Use the Mobile-Friendly Test or the Rich Results Test from Google to manually validate your key pages. These unit tests are not subject to sampling and provide a reliable vision page by page.

• Deploy a third-party crawler (Screaming Frog, OnCrawl, Botify) for a comprehensive quarterly audit
• Identify and manually monitor high-value strategic pages
• Never conclude based solely on Search Console’s aggregated reports
• Compare the volume of pages in the index coverage report with that of the technical reports
• Use Google’s unit testing tools (mobile, rich results) for priority pages
• Consistently check organic traffic and rankings before diagnosing a sampling issue