Google Search Console: The 6 Reports That Actually Tell You Something

Verification, and which property type

Two property types, and the choice matters more than it looks.

• Domain property — verified by DNS record. Covers every subdomain and both protocols. This is what you want.
• URL-prefix property — verified by file, tag or analytics. Covers only that exact prefix, so https://example.com and https://www.example.com are separate properties with separate data.

Set up a domain property if you can add a DNS record. Sites that skip this end up with their traffic split across three or four properties and no complete picture of anything.

One thing worth doing on day one: data starts accumulating from verification, and it is not backfilled. If you have not verified a property yet, doing it now is the cheapest thing on this list.

1. Performance — and how it misleads

Four metrics, and one of them causes most of the wrong conclusions people draw from this tool.

Impressions — how often a URL appeared in results. Note that appearing on page seven counts as an impression. Rising impressions can mean growing visibility or simply that you now appear, badly, for more things.

Clicks — actual clicks through to your site. The honest number.

CTR — clicks divided by impressions. Useful, with the caveat that it drops naturally as you gain impressions for queries you rank poorly for. A falling CTR alongside rising impressions is often fine.

Average position — the one to be careful with. It is the average of your highest position across every impression, aggregated over whatever you have filtered. So:

• An unfiltered site-wide average position is close to meaningless — it blends your brand terms with everything you barely rank for.
• It can worsen because you started appearing for new, harder queries. That is growth showing up as a decline.
• It can improve because you stopped appearing for queries you ranked badly for. That is loss showing up as progress.

Average position is a trend indicator for a single query on a single page. Anything more aggregated than that, and you are averaging things that should not be averaged.

The filters are where the value is. Filter by query, by page, by country, by device, and compare periods. And prefer year on year over month on month for anything seasonal — comparing December to November tells you about Christmas, not about your SEO.

2. Queries by page

This is the single most useful view in the tool and it is slightly buried. In Performance, filter to one page, then open the Queries tab. You now see exactly what that URL gets found for.

What it tells you that nothing else does:

• Whether the page is understood correctly. If your Meta Ads page surfaces mostly for unrelated queries, the page is not saying what you think it says.
• Near-miss opportunities. Queries at positions 5–15 with real impressions. These are pages Google already considers relevant — usually the cheapest wins available, and often just a matter of covering the query properly on the page.
• Cannibalisation. Filter by query instead, and if several of your URLs appear for it, they are competing. Consolidate, or differentiate them properly.
• Entity and brand health. Filter queries containing your name or brand. Rising branded impressions, and the right page answering them, is the clearest read on whether your identity signals are landing.

A reasonable monthly habit: pull the top twenty pages, look at each one’s query list, and note anything sitting at position 5–15 with meaningful impressions. That list is your content plan, derived from what Google already believes about you rather than from a keyword tool’s guess.

3. Pages: what the statuses mean

The Pages report splits your URLs into indexed and not indexed, with reasons. The reasons are where people misread things badly.

Two things to internalise. “Crawled — currently not indexed” is not a bug you can configure away. Requesting indexing repeatedly does not fix it; it is Google saying the page is not worth a slot. The fix is the page. And a large not-indexed count is not automatically bad — tag pages, paginated archives and filtered URLs being excluded is usually correct. Look at whether the right pages are indexed, not at the ratio.

4. URL Inspection

The diagnostic tool for one URL, and the closest thing to a direct answer you get.

It tells you whether the URL is indexed, when it was last crawled, which canonical Google selected versus which you declared, whether it was discovered via sitemap or link, and how the page rendered — including the HTML Google actually saw.

Three genuinely valuable uses:

1. Compare canonicals. “User-declared” versus “Google-selected”. When they differ, you have a signal conflict worth chasing.
2. Check rendered HTML. Use “Test live URL” and view the rendered source. For any JavaScript-heavy page this is how you confirm your content and schema actually exist as far as Google is concerned.
3. Request indexing after a genuine change. Useful for a handful of important URLs; not a lever for pushing hundreds of pages, and it does not override quality judgements.

5. Sitemaps

Submit your sitemap here and check two things: that Google read it without errors, and that the URL count it discovered roughly matches what you expect.

A discovered count well below your actual page count usually means the sitemap is stale or generated from an incomplete source. Well above usually means it includes URLs that should not be there — redirects, noindexed pages, parameter variants.

Worth remembering that a sitemap is a suggestion. It helps discovery; it does not compel indexing, and adding a page to it will not overcome a quality judgement.

6. Enhancements and Core Web Vitals

The Enhancements section reports structured data Google found and any errors in it, per type. This is the only place you see schema errors across your whole site rather than one URL at a time — which makes it the right place to catch a template that broke and stripped markup from a thousand pages.

Core Web Vitals here is based on real user data from the Chrome UX Report, which makes it meaningfully different from a lab tool like Lighthouse. Lab tools tell you what is slow on a simulated device; this tells you what your actual visitors experienced. When they disagree, trust this one for prioritisation and the lab tool for diagnosis.

Note that CWR data needs sufficient traffic to report, so low-traffic sites may see nothing here at all. That is not an error.

What to largely ignore

• Site-wide average position. Discussed above. Not a KPI.
• The Links report’s totals. Directionally interesting, substantially incomplete. Do not use it for link auditing.
• Small day-to-day fluctuations. Data is sampled, delayed by two to three days, and noisy. Look at weeks.
• The anonymised query gap. Your query totals will not sum to your click totals, because rare queries are withheld for privacy. This is expected and unfixable.
• Removals. A temporary emergency tool, not part of normal SEO work.

A monthly routine

That last point catches people out. There is no archive, no recovery, and no way to get 2024 back once it has rolled off. A monthly CSV export into a spreadsheet takes two minutes and is the only way to have multi-year data later.

FAQ