Definition
Google Search Console is a free service, described in its help centre as a way to monitor, maintain and troubleshoot a site's presence in Google Search results. Access requires verifying ownership of a property, either a domain property covering every protocol and subdomain, or a URL-prefix property covering one exact prefix. Once verified, the Performance report shows queries, clicks, impressions, click-through rate and average position for the site in Google Search, and in Discover and News where the site is eligible; the indexing reports show which URLs are indexed and which are excluded and why; and the experience reports show Core Web Vitals and HTTPS status. The registry one-liner is the short version: queries, clicks, impressions, CTR, position and index coverage for a verified site. The Search Console guide covers each report in working detail, and the Search Console expertise page sets out how it is used in SEO engagements.
How it works
Search Console compiles data from Google's own systems rather than from a tag on the site, which is why no code is installed and why its figures do not match analytics tools. An impression is counted when a link to the site appears in a results page a user viewed; a click is a click on that link; position is the average ranking position of the site's link across those impressions, with Google documenting that it records the topmost position when the site appears more than once on a page. The Performance report can be filtered and compared by query, page, country, device, search appearance and date, and the help centre notes that some queries are omitted from the query view to protect privacy, so the query table totals less than the overall total. Data is retained for a documented period.
| Report | Question it answers |
|---|---|
| Performance | Which queries and pages earned impressions and clicks, at what position |
| Pages (indexing) | Which URLs are indexed, which are not, and the reason for each |
| URL Inspection | What Google knows about one URL, with a live test and a request to index |
| Sitemaps | Which sitemaps were submitted and read, and how many URLs they listed |
| Core Web Vitals and HTTPS | Which URL groups fail the page experience criteria |
| Manual actions and Security issues | Whether the site is penalised or flagged |
| Links | Which sites link in, which pages are linked, and internal link counts |
| Crawl stats (settings) | How often and how successfully Googlebot fetched the site |
Indexing reports classify each known URL with a reason: crawled but not indexed, discovered but not crawled,
excluded by noindex, duplicate with a different canonical, page with redirect, and so on. The
crawling and indexing guide explains what each reason implies for the
site. Search Console also sends email alerts for new issues and can be linked to Google Analytics so that query
data appears alongside behaviour data.
In practice
Day to day, Search Console answers two questions: what is the site earning from search, and is anything
stopping pages from being found. The first is read from the Performance report with a comparison period and a
filter on page or query group; the second from the Pages report and URL Inspection. Both become more useful
when the site is a domain property, so that www, bare-domain and http variants are all included, and when
sitemaps are submitted so that the indexing report can compare what the site intends to have indexed with what
Google has indexed. Exports to a spreadsheet or to Looker Studio are the usual route to a longer history.
Diagnosing an organic traffic fall
A publisher sees organic sessions down in its analytics. In Search Console it compares the last 28 days with
the previous 28: clicks fell from 61,400 to 47,900, impressions were roughly flat at 2.1 million, and average
position moved from 8.4 to 11.2. Filtering by page shows the loss concentrated on 35 recipe URLs, and the Pages
report lists 22 of them as "Crawled - currently not indexed" since a template change. URL Inspection on one of
them shows the live page returning a noindex header injected by a caching plugin. Removing the header and
requesting indexing restores the pages within a fortnight, and clicks recover to 59,800 in the following period.
What it is not
Search Console is not an analytics platform: it knows nothing about what happened after the click, has no sessions or conversions, and its click counts differ from analytics sessions for documented reasons such as cached pages and blocked scripts. It is not a rank tracker: average position is an average over the impressions Google chose to show, not a measurement of a fixed query at a fixed location. It is not a ranking lever; nothing submitted through it, including a sitemap or an indexing request, improves position. It is not a crawler log, although the crawl stats report summarises Googlebot activity. And it covers Google only; other engines have their own equivalents, and advertising data lives in Google Ads, which shows paid clicks on the same queries from a different system.
Related terms
Search Console is the measurement surface for SEOImproving visibility in unpaid search results., and its Performance report is where most organic reporting starts. Its search appearance filter shows which result features (rich results, videos, product snippets) the site earned on the SERPThe page of results a search engine shows., which is often the explanation for a CTR change at a stable position. The crawl stats report and the indexing reasons are the evidence for decisions about Crawl BudgetHow much a search engine crawls a site. on large sites: a long tail of "discovered, currently not indexed" URLs is usually a signal that the site exposes more URLs than Google is willing to fetch. Sitemaps, robots rules, canonical tags and structured data all show their effects here, which is why the tool sits at the end of every technical SEO change and at the start of every organic report.
Common mistakes
- Verifying a URL-prefix property and missing half the site. A domain property includes every subdomain and protocol; a prefix property does not.
- Reading average position as a rank. It is an average across every impression, including low positions in image and video results.
- Comparing Search Console clicks with analytics sessions. They are defined and collected differently; use each for what it measures.
- Treating "crawled, currently not indexed" as a bug. For thin or duplicate pages it is Google's verdict on the content.
- Ignoring the omitted queries. The query table never sums to the total; the gap is by design.
- Exporting without the date comparison. A single period hides the trend that explains the number.
- Requesting indexing as a fix. The request asks Google to look again; if the page was excluded for a reason, the reason has to change first.
