Diagnosing a loss of visibility
It separates a drop caused by deindexing, by falling positions, or by falling demand — three causes with opposite remedies.
Google Search Console is the free service through which Google shows a site owner how their domain is crawled, indexed and displayed in search. It provides the real queries, impressions and clicks, indexing status page by page, technical issues detected and field data for Core Web Vitals.
In one sentence
Search Console shows what Google actually sees of your site — and what is stopping it from getting in.
Key points
Term at a glance
Search Console does not measure your site: it measures the relationship between your site and the search engine. That distinction is practical, because it explains why its figures never quite match an audience analytics tool — they are not counting the same thing, from the same side.
The performance report gives four measures for each query, page, country and device: impressions, clicks, click-through rate and average position. Average position is an average of appearances, not a stable ranking: a page seen at position 3 on one query and 40 on another will show an average matching no observable reality.
The indexing report is the most decisive and the least consulted. It sorts known URLs into indexed and not indexed, and gives a precise reason for each exclusion: blocked by robots.txt, marked noindex, alternate page with proper canonical tag, crawled but not indexed, discovered but not crawled. Those reasons point to very different fixes.
The URL inspection tool completes the picture by showing, for one address, the last crawl, the rendering Google obtained, the canonical it settled on and the structured data detected. It is the only way to see the page as Google rebuilds it, with JavaScript executed.
The domain property, validated through a DNS record, covers every subdomain and http/https variant. It removes the need to compare three partial properties.
The XML sitemap helps Google discover URLs and, above all, lets you compare what you declare against what is actually indexed. The gap between the two is a diagnostic in itself.
Start with the reasons that signal a genuine problem — crawled but not indexed, discovered but not crawled — rather than with deliberate, normal exclusions.
Group the queries shown by intent rather than by volume: they reveal the questions your site already answers, and those it attracts without answering.
Joining them lets you follow a query through to conversion. Above all, check that the two tools are not read as though they measured the same thing.
History stops at 16 months. A monthly export, or a BigQuery connection through bulk export, preserves the long series needed for year-on-year comparison.
A distributor in the Laurentians sees organic traffic drop 30% over a quarter. GA4 confirms the fall but gives no cause. The Search Console indexing report shows 1,200 product pages moving to "crawled, currently not indexed" on the same date. URL inspection reveals a canonical tag pointing every size variant at a deleted page, introduced during a catalogue update. The fix is one line of code; without Search Console, the diagnosis would have taken weeks.
It separates a drop caused by deindexing, by falling positions, or by falling demand — three causes with opposite remedies.
URL inspection confirms crawling, rendering and indexing without waiting for the page to appear in results.
Queries with high impressions and low click-through rate flag questions the site is judged partly relevant to.
The report is based on field data from Chrome visitors, not on an isolated lab test.
| Google Search Console | Google Analytics 4 | |
|---|---|---|
| Measurement point | The search engine, before the click | The site, after the visitor arrives |
| Key data | Queries, impressions, position | Sessions, events, conversions |
| Coverage | Google only | Every traffic source |
| Consent | Not applicable, engine-side data | Subject to the visitor's consent |
Without Search Console, a drop in organic traffic stays a fact with no cause: you watch the number fall without knowing whether to fix a technical issue, produce content, or do nothing because demand simply receded. It is also the only place where a page technically invisible to Google announces itself as such — an entire site can be beautifully written and stay absent from results for a reason that comes down to three lines of configuration. For an SME it is the best value-for-cost tool in digital marketing: it is free.
They measure two distinct moments. Search Console counts clicks leaving the engine; GA4 counts sessions arriving on the site, and only if tracking loads and consent allows it. A 10 to 20% gap is normal.
From a few hours to several weeks, depending on how often the site is crawled. Requesting indexing through URL inspection speeds up discovery but does not guarantee indexing.
Google saw the page and chose not to keep it. It is usually a quality or redundancy signal: thin content, very close to another page, or with no identifiable search demand.
No, if you use a DNS-validated domain property: it covers everything. URL-prefix properties remain useful for isolating the analysis of one specific section.
For diagnosing your own site it is often sufficient and more reliable. It says nothing about competitors, global search volume or backlinks — that is where third-party tools keep their value.
Watching organic traffic fall without knowing whether the cause is technical, editorial or competitive?
Get your visibility diagnosed