Before you start
- Access to the site sufficient to verify ownership: a DNS record (which verifies the whole domain), an HTML file or tag on the site, or an existing analytics or tag manager account.
- A list of the site's important pages and queries from the keyword map, so that the reports can be read against what matters.
- The XML sitemap URL and the robots.txt.
- A place to record exports and annotations — a spreadsheet at minimum, since history is limited.
Create the property and verify it
Add a Domain property (covers every protocol and subdomain, verified by DNS) or a URL-prefix property (one exact prefix, verified by file, tag or other methods). Domain properties are the better default; URL-prefix properties are useful for separating a section or a subdomain. Verify, then add the team as users with the right permission level. Verification that lapses (a removed tag, a changed DNS record) silently ends data access, so use a method that will survive site changes.
Submit the sitemap and check robots
Under Sitemaps, submit the sitemap or sitemap index URL and confirm it is read without errors; the report shows discovered URL counts per file. Check the robots.txt report shows the current file and no fetch errors. Neither makes pages rank; both make the rest of the data interpretable, because the engine now knows which URLs the site considers canonical.
Read the performance report correctly
The Search results performance report shows clicks, impressions, click-through rate (CTRShare of impressions that result in a click.) and average positionMean ranking of a site's topmost result across the impressions it received., filterable by query, page, country, device, search appearance and date, with comparison between periods. The rules for reading it:
- Always filter. Site totals mix brand and non-brand, every country and every device, which hides everything. Filter to non-brand queries (exclude the brand terms), to the main country, and to a page group or query group.
- Read impressions with position. Impressions rise when the site starts appearing lower on more results pages, which drags average position down while visibility improves; that is normal and good.
- Read CTR against position and against result features. A query answered in a snippet or crowded with ads has a lower click-through at the same position; comparing CTR across queries without that context is meaningless.
- Compare like periods. Weeks against weeks, the same weekdays, seasonally comparable months.
- Expect sampling. The long tailA specific, usually longer query searched rarely on its own but often as a group. is anonymised and the table is capped; export through the API or bulk export for complete data.
- Use the page-and-query view for cannibalisation: filter to a query and see which pages received impressions, as the cannibalisation guide describes.
Read the page indexing report
The Page Indexing report counts indexed and not-indexed pages and lists the reasons for the latter: not found, redirects, blocked by robots, excluded by noindex, duplicates with and without a chosen canonical, crawled but not indexed, discovered but not crawled, soft 404sA page that returns a success status code but shows an error, empty or not-found message., server errors. Each reason is a different problem with a different fix, covered in the crawling and indexing guide. What matters is the trend per reason and whether the pages listed are ones that should be indexed. A large "not indexed" count is normal on a site with parameters and pagination; an important page in the list is not. Validate fixes through the report once deployed and watch the validation complete.
Read the experience and enhancement reports
Core Web Vitals (Core Web VitalsGoogle's user experience metrics.) shows URL groups rated good, needs improvement or poor from field data, for mobile and desktop; it is grouped by template behaviour, so a single fix moves many URLs, and it lags fixes by weeks. HTTPS reports pages served insecurely. The enhancement reports list structured data types found and their errors, per type — the place to see whether product, article, breadcrumb or FAQ markup is being read. The Core Web Vitals guide covers the metrics.
Read the links report
Links lists the top externally linked pages, top linking sites and top anchor text as Google sees them, and the internal link counts per page. It is a sample and lags, but it is Google's own view, which no third-party tool has. Use it to confirm that important pages receive internal links, to spot unexpected linking sites, and as one input to the link audit.
Use URL inspection for individual pages
URL Inspection shows, for one URL, whether it is indexed, the canonical Google chose (Canonical TagTells search engines the preferred URL.) versus the one declared, when it was last crawled, whether it is mobile-usable, and which structured data was detected; a live test fetches and renders the page now, showing the rendered HTML and screenshot. It is the fastest way to answer "why is this page not showing" and to confirm a fix on one URL before the report catches up. Requesting indexing after a fix nudges a recrawl; it is not a guarantee and not a bulk tool. When a whole new site cannot be found, the page on a website not showing on Google starts from this inspection and works through the blocks and gaps in order.
Check manual actions and security issues
Both reports should read "no issues". A manual action names the policy violated and the affected pages, and must be fixed and submitted for reconsideration. A security issue flags hacked content or malware. Make sure email notifications are on for the property, because these are the two reports that cannot wait for the monthly review.
Set up exports and annotations
Because history is limited and tables are capped, export monthly, or connect the bulk export to a data warehouse. Keep an annotations log: every site change, content publish, migration, campaign and known Google update with its date, so that changes in the reports can be read against causes. Annotation is the step most teams skip and most regret.
Establish the routine
Weekly: page indexing trend and any important page newly excluded; Core Web Vitals status; manual actions and security; performance for priority clusters against last week; new structured data errors. Monthly: the full KPI chain from the KPIs guide with annotations; sitemap and links checks; export. After any deployment: URL inspection on affected templates and the indexing report over the following weeks. The technical audit tutorial uses these reports as one of its three data views.
Reading a drop correctly
A site's monthly clicks fall. Unfiltered, the chart shows a decline and the team assumes a ranking problem. Filtered to non-brand queries in the main country, clicks are flat; the decline is brand clicks, and the annotations log shows a paid brand campaign ended that month, which had been driving brand searches. Filtered to the product page group, impressions have risen while average position has dropped slightly: the pages are appearing on more results pages (SERPThe page of results a search engine shows.) at lower positions, which is growth. One query in the priority set does show a real drop, and the page-and-query view shows two URLs now sharing it — a new blog post has begun competing with the service page. The actions are a consolidation, not a site-wide investigation.
What Search Console does not tell you
- Conversions and revenue. That is analytics, joined by landing page.
- Competitors' performance. That is rank tracking and backlink tools.
- Why a ranking changed. It reports the change; the annotations log and the audit find the cause.
- Daily positions per query by location. Rank tracking.
- The complete long tail. Exports and the bulk export get closer; some anonymisation remains.
The rank tracking guide explains how a tracker's sampled positions for chosen queries, locations and devices differ from this report's average position, and how to read the two together.
After a redesign or domain change, the reports above are the starting point for recovering organic traffic after a migration.
When the enhancement reports look clean but nothing appears in results, the page on rich results that are not showing separates errors from eligibility and Google's discretion.
To turn performance report clicks and impressions into a verdict, follow the organic search CTR calculator tutorial, which sets the target from ranking position.
When the aim is more visits rather than more leads or sales, organic search for website traffic shows how the performance report's clicks, click-through and indexing data become a traffic KPI set.
When the Page Indexing report shows blocked URLs that are indexed anyway, the diagnosis on indexed, though blocked by robots.txt traces each group to the rule or link behind it.
Page-level clicks and positions from this report are the raw material for the page-group split design for title tag tests, which compares a changed group of pages with an unchanged one over the same weeks.
When Search Console clicks and analytics sessions disagree, the comparison of Search Console and GA4 explains which questions each tool answers and why their counts never match.
Verify it worked
- Open the property and its users. Data should be flowing, with a verification method that will survive site changes.
- Open Sitemaps and the robots.txt report. The sitemap should be read without errors, and robots.txt should show the current file with no fetch errors.
- Filter the performance report to non-brand queries, the main country and one page group. The figures should differ from the unfiltered site totals.
- Open the page indexing report. Every important page should be indexed, and each not-indexed reason should be one you expect.
- Inspect one key URL. It should show as indexed, with the canonical Google chose matching the one you declared.
If it goes wrong
- Data stops or the property loses access. Verification has probably lapsed, through a removed tag or a changed DNS record. Verify again by a method that survives site changes.
- Clicks fall in the performance report. Filter before concluding. Split brand from non-brand, then country and device, and read the result against the annotations log.
- An important page is not indexed. Inspect the URL and match its reason in the indexing report to the fix. Validate the fix and watch the validation complete.
- Core Web Vitals look unchanged after a fix. The report lags fixes by weeks, so check again later before changing anything else.
- Search Console clicks differ from analytics sessions. They count different events, so neither is wrong. Do not force them to match.
Next steps
- Read the SEO KPIs and reporting guide to build the monthly KPI chain from these reports.
- Read the crawling and indexing guide to fix each not-indexed reason in the page indexing report.
- Read the technical SEO audit tutorial to see these reports used as one of its three data views.
Common mistakes
- Reading site totals. Brand, country and device mixed together.
- Treating average position as a rank. It is a mean across impressions.
- Comparing CTR across queries with different result features.
- Ignoring the indexing report until something breaks.
- No annotations. Every change becomes a mystery.
- Letting verification lapse. Months of data lost.
- Requesting indexing as a strategy. A nudge for one URL, not a fix.
The measurement hub places Search Console among the other sources; the Google Search Console expertise page describes the author's working use of the reports.
