Key takeaways
- The Pages report is the most valuable and least used report in SEO.
- Position data is an average across many searches — treat it as a trend, not a rank tracker.
- "Crawled – currently not indexed" is a quality signal; more internal links will not fix it.
- Filter by position 8–20 to find your fastest available wins.
Google Search Console is the only place Google tells you directly what it thinks of your site — which pages it has indexed, which it has rejected and why, and what searches it shows you for. It is free, and most businesses either never set it up or look only at the traffic graph.
1. The Pages report — start here
Found under Indexing → Pages. It splits every URL Google knows about into indexed and not indexed, with a reason for each exclusion. It is the most useful report in the tool and the one people skip.
The first thing to check is simple arithmetic: how many pages have you published, and how many are indexed? A large gap is almost always your biggest SEO problem.
| Status | What it means | What to do |
|---|---|---|
| Crawled – currently not indexed | Google fetched the page and decided it was not worth indexing. A quality judgement. | Make the page materially better or consolidate it into a stronger page. More links will not fix this. |
| Discovered – currently not indexed | Google knows the URL exists but has not prioritised crawling it. | Strengthen internal links to it and reduce low-value URLs competing for crawl budget. |
| Duplicate without user-selected canonical | Multiple URLs serve near-identical content and none declares a canonical. | Add self-referencing canonicals and fix the URL variants causing it. |
| Alternate page with proper canonical tag | Working as intended. | Nothing — unless the canonical target is wrong. |
| Soft 404 | Returns HTTP 200 but looks empty or error-like to Google. | Return a real 404, or add genuine content if the page should exist. |
| Excluded by "noindex" tag | A noindex directive is present. | Confirm it is intentional. Stray noindex tags after a staging launch are extremely common. |
| Blocked by robots.txt | A disallow rule prevents crawling. | Confirm intentional. If the page should rank, remove the rule. |
2. The Performance report — queries
Performance → Search results shows the actual searches your site appears for. Four metrics: clicks, impressions, click-through rate and average position. All four are toggled on and off at the top, and by default position is switched off — turn it on.
- Sort by impressions, not clicks. Impressions show where Google thinks you are relevant, which is where the opportunity is
- Filter by position 8–20. These are your fastest wins — proven relevance, just off page one
- Find high impressions with low CTR. Either the position is too low or the title is not earning the click
- Look for queries you never targeted. These frequently reveal demand you did not know about
- Compare date ranges. A three-month period against the previous three months shows direction, which a single snapshot cannot
3. The Pages tab within Performance
Switch from Queries to Pages and you get per-URL performance. This is where cannibalisation becomes visible: filter by a specific query and look at how many of your URLs receive impressions for it. If several do, they are competing with each other.
It is also how you diagnose a drop. If site-wide traffic fell, check whether it fell across all pages or in one section. A section-level drop points at a technical or template change; a site-wide one usually points at an algorithm update or something that affected the whole domain.
4. Core Web Vitals
Under Experience → Core Web Vitals. This is field data from real Chrome users, which is what Google actually uses. It is far more meaningful than a Lighthouse score, and the two often disagree.
- URLs are grouped into Good, Needs improvement and Poor, and grouped by similar page templates
- Fixing one template usually fixes every URL in that group at once
- Data lags by roughly 28 days, so improvements take about a month to appear
- A "not enough data" state simply means low traffic, not a problem
The URL Inspection tool
Paste any URL into the search bar at the top. It tells you whether the page is indexed, when it was last crawled, which canonical Google selected, and — under "View crawled page" — the rendered HTML the crawler actually received.
That last part is the fastest way to diagnose a JavaScript rendering problem. If your content is missing from the rendered HTML, no amount of on-page optimisation will help until that is fixed.
What to mostly ignore
- Links report — the data is incomplete and heavily sampled. Useful for spotting something dramatic, not for link analysis
- Request indexing, repeatedly — one request per page is enough. Repeated requests do not accelerate anything
- Day-to-day position fluctuation — noise. Compare periods of a month or more
- Removals tool — for genuine emergencies only, and the effect is temporary
- Enhancement reports for schema you did not intend — validate what you added, ignore incidental noise
A useful monthly routine
- 01
Check the Pages report
Has indexed page count moved? Any new exclusion category growing?
- 02
Compare three months against the previous three
Impressions, clicks and average position. Direction, not snapshots.
- 03
Filter queries to positions 8–20
Pick two or three to improve this month.
- 04
Check for new queries
Anything appearing that suggests a new opportunity?
- 05
Review page-level movement
Which URLs gained and which lost, and does anything correlate with a release?
- 06
Glance at Core Web Vitals
Only if a template changed or a new script was added.
Twenty minutes a month with those six steps tells you more about your SEO than most agency reports do. If the numbers are not moving, that is covered in why your website is not ranking.