Search Console has more reports than anyone needs, and five of them answer almost every question worth asking. One of those five did not exist three months ago.
The others are not useless — they are situational, and treating them as a routine checklist is how people spend an hour a week learning nothing. This guide covers what each of the five actually tells you, the specific reading that matters in each, and what to skip.
1. Performance — but read as a comparison
The question it answers: is search sending more or fewer people, and why?
Most people read the four totals at the top and stop. The totals are the least informative thing on the page.
What to read instead:
- Impressions against clicks. Impressions up with clicks flat means you are appearing and not being chosen — a title, description or SERP-feature problem, not a ranking problem.
- Queries, sorted by impressions rather than clicks. High-impression, low-click queries are where the cheapest wins live.
- Pages, compared period on period. One page losing badly is a different problem from everything drifting down.
- The comparison tab, always. A number without a previous number is not information.
One thing to know about position: average position is an average of a distribution, so it can improve while traffic falls, and vice versa. Treat it as directional and never as a KPI.
2. Page indexing — read the reasons, not the total
The question it answers: can Google store your pages, and if not, why not?
The total is close to meaningless. Most sites carry a large "not indexed" number made up entirely of statuses that are working as intended — alternate pages with canonicals, redirects, 404s, pages you marked noindex.
What to read: which reason is growing. A category that was 40 last month and is 4,000 this month is the signal, whatever the total.
Two reasons deserve immediate attention:
- "Indexed, though blocked by robots.txt" — a page is in the index that you told Google not to crawl, and Google could not read it before indexing it
- Server error (5xx) — anything above a trickle means Google is meeting a struggling server
Every status decoded, and which need no action at all, is in why isn't my page indexed.
3. URL Inspection — the most useful tool in the product
The question it answers: what does Google actually think about this specific URL?
For one URL it gives you indexing status, the canonical Google selected — which may not be yours — crawl date, discovery route, and whether it can fetch and render the live page.
Use "Test live URL" rather than the stored result. The stored result reflects the last crawl, which may predate the fix you just deployed. This single habit prevents most "I fixed it and nothing changed" confusion.
The most valuable line in it is "Google-selected canonical." When it differs from yours, that is usually the whole explanation for a page not performing, and nothing else in Search Console tells you as directly.
4. Core Web Vitals — field data, not lab scores
The question it answers: are real visitors getting a fast, stable page?
This report uses field data from actual Chrome users, which is what Google acts on. A Lighthouse score is a lab simulation on assumed hardware and is not the same measurement — sites regularly score 100 in Lighthouse and fail here.
The thresholds:
| Metric | Good |
|---|---|
| Largest Contentful Paint | Under 2.5 seconds |
| Interaction to Next Paint | Under 200 milliseconds |
| Cumulative Layout Shift | Under 0.1 |
Two practical notes. The report groups URLs by template, so fixing one page usually fixes a whole group — work at template level. And it needs traffic before it shows anything; a low-traffic site may have no data at all, which is not a fault.
Read this one monthly rather than weekly. Field data accumulates over a 28-day window, so week-to-week movement is mostly noise.
5. AI performance — the newest report, and the least understood
The question it answers: are you appearing in Google's AI answers?
This is the one that did not exist recently. Google began breaking out AI performance in the UK on 3 June 2026 and completed the global rollout on 31 August 2026.
What it gives you: impressions for AI Overviews, AI Mode and generative AI in Discover, separated from ordinary Search, filterable by page, country, device and date.
What it does not give you:
- No clicks, for any AI surface
- No click-through rate, because there are no clicks to divide by
- No query data at any level
- No per-link position — every link inside an AI Overview shares one position slot
So it is a visibility report, not a traffic report. You can see that you appeared. You cannot see whether anyone came.
The reading that matters: AI impressions rising while overall Search clicks for the same pages stay flat. That is being cited without being visited, and until August there was no way to observe it.
One thing to understand about the numbers: this is a breakout of data that was always inside your Search Console totals, not new data. Your historical figures did not change — those impressions were simply never itemised.
The measurement picture around this, including what GA4 can and cannot add, is in how to track AI search traffic.
What to skip
Links. Interesting occasionally, actionable rarely. It reports what Google has found, undercounts consistently, and there is nothing to do with most of it.
Enhancements reports for types you do not use. They appear based on markup Google detects. If a report exists for a type you did not intend to publish, that is worth investigating — otherwise ignore it. What a blog actually needs is in schema markup for blogs.
Removals. A genuine emergency tool for getting something out of results within hours. Not a routine report, and not a substitute for noindex — removals expire.
Crawl stats. Buried in Settings, occasionally invaluable when diagnosing server problems under crawl load, irrelevant the rest of the time.
Security and manual actions. Check they are empty. If they are not, that outranks everything else on this page.
A workable routine
Weekly, ten minutes: Performance with the comparison on, and Page indexing reasons for anything growing.
Monthly, twenty minutes: Core Web Vitals field data, AI performance impressions against Search clicks, and a URL Inspection on any page that matters and looks wrong.
After every deployment, migration or theme change: URL Inspection with live test on three important pages, and a look at robots.txt. Those two checks catch the failures that otherwise go unnoticed for weeks — the mechanism is in robots.txt mistakes.
Set the property up properly once. Use a Domain property rather than a URL-prefix property where you can, so subdomains and both protocols are covered in one place. Sites frequently discover months of missing data because they verified https://www. and their traffic went to the apex.
Frequently asked questions
Which Search Console reports actually matter?
Performance, Page indexing, URL Inspection, Core Web Vitals, and the AI performance report added in 2026. Between them they answer whether search is sending traffic, whether Google can store your pages, what Google thinks of a specific URL, whether real visitors get a fast page, and whether you appear in AI answers.
Does Search Console show AI Overviews data?
Yes, since 31 August 2026 globally — impressions for AI Overviews, AI Mode and generative AI in Discover, broken out from ordinary Search and filterable by page, country and device. It shows no clicks, no click-through rate and no query data for those surfaces.
Why is my "not indexed" number so high?
Usually because most of it is working as intended — alternate pages with canonicals, redirects, 404s and pages you marked noindex all count toward it. The reasons are the data; the total is not. Watch which reason is growing rather than the headline figure.
What is Google-selected canonical?
The URL Google chose to represent a page, which may differ from the canonical you set. When it differs, that is usually the explanation for a page not performing, and URL Inspection is the only place that tells you directly.
Should I use Lighthouse or Search Console for Core Web Vitals?
Search Console, because it reports field data from real Chrome users, which is what Google acts on. Lighthouse is a lab simulation on assumed hardware and connection — sites routinely score 100 there while failing Core Web Vitals in the field.
How often should I check Search Console?
Ten minutes weekly for Performance and Page indexing reasons, twenty minutes monthly for Core Web Vitals and AI performance, plus URL Inspection after any deployment or migration. Core Web Vitals uses a 28-day window, so weekly readings of it are mostly noise.
Why does my average position improve while traffic falls?
Average position is an average across a distribution of queries and impressions, so it can move independently of clicks — losing visibility on high-volume queries while gaining on low-volume ones improves the average and reduces traffic. Treat it as directional, never as a KPI.
Should I use a domain property or a URL-prefix property?
A Domain property where possible, because it covers all subdomains and both protocols in one place. Sites often discover months of missing data because they verified one prefix while traffic arrived on another.
What to do next
Open Performance, turn on the comparison against the previous period, and look at impressions against clicks rather than the totals. If impressions are up and clicks are flat, your problem is presentation rather than ranking — and that is a different and much cheaper fix.
Then check whether your AI performance report has data. It only completed global rollout on 31 August 2026, so there may be less history than you expect.
Related guides
- Technical SEO: what actually breaks rankings — the audit order these reports serve
- Why isn't my page indexed? A diagnostic — every Page Indexing status decoded
- robots.txt mistakes that de-index whole sites — the check to run after every deployment
- Schema markup for blogs: what to add and how — what the Enhancements reports are reporting on
- How to track AI search traffic in GA4 — what analytics adds to the AI picture
Free: The 60-Minute Email Authentication Fix
A no-fluff checklist to set up SPF, DKIM & DMARC correctly and pass Gmail & Yahoo's sender requirements.
Muhammad Basim has worked in digital marketing since 2013, focused on email deliverability and AI-assisted content production. He is the author of The Email Deliverability Playbook and The Email Copywriting Playbook.
Related Articles
Email Deliverability: Why Authenticated Emails Still Land in Spam
Email deliverability is whether your message reaches the inbox rather than the spam folder, and it is decided by three factors: technical infrastructure, list quality, and sending behaviour. Authentication is one item inside the first factor. Getting it right is necessary, and it is nowhere near sufficient. That gap explains the most common complaint in […]
Linkable Assets: What Actually Earns Links
A link is a citation, and a citation requires that someone writing about your subject needed you to make their point. That is the whole mechanism, and it is the reason most "linkable content" earns nothing. The test, before you build anything: Could a writer covering this topic finish their sentence without referencing you? If […]
Link Outreach Emails That Get Replies
Outreach fails for exactly two reasons, and only one of them gets written about. The first is that the email was not worth replying to. That is the reason every guide addresses, and the advice — personalise, be brief, offer value — is correct and insufficient. The second is that the email never arrived. Nobody […]