Search Console is the only place you see Google's own numbers for your site — what people searched, what they clicked, and which of your pages Google actually indexed. This course gets you from zero to a working weekly routine.
⏱ 4 min read
Google Search Console (GSC) is a free tool that reports how your site behaves inside Google Search. It is Google's own data about your site: which queries showed your pages, how often people clicked, which URLs are indexed, and which are not. No third-party tool can see those numbers — they estimate; Search Console measures.
It is free for anyone who can prove they own a site. There are no paid tiers and no premium reports. If you build or manage websites, there is no reason not to have every site verified.
Search Console is not analytics. It stops at the click. It tells you someone searched "brecksville web design", saw your page, and clicked it — then hands off. What that person did next (read three pages, filled a form, left in four seconds) lives in Google Analytics. The two tools are complements, not substitutes, and the course on GA4 picks up exactly where this one stops.
It is also not a ranking tool. The average position it reports is an average across many searches, locations, and devices. Treat it as a trend line, not a scoreboard.
Two quirks catch every beginner. First, there is a processing delay of roughly two to three days, so today's data is never there yet. Second, Google filters out very low-volume queries for privacy, so the query table will never add up exactly to the chart totals. Both are normal. Reports also occasionally stall — the Page Indexing report ran weeks behind during the summer of 2026 before Google caught it up. When a report looks frozen, check its "last updated" stamp before assuming your site broke.
Intro to Google Search Console — Google Search Central. Daniel Waisberg's five-minute tour of the interface — the fastest orientation there is. Open on YouTube ↗
Two permission levels cover almost every situation. Owners can add and remove users, change settings, and delete the property. Full users can see all data and use most tools but cannot manage access. Restricted users see most reports and can change nothing. For agency work, the cleanest arrangement is for the client to own the property and grant the agency full access — the client keeps control when the relationship ends, and nobody shares a password to do it. Verify the site under your own account as well if you can; two verified owners means losing one does not lock anyone out.
⏱ 4 min read
A property is one site (or one slice of a site) inside Search Console. Your first decision is which kind to create, and it is worth ten seconds of thought because it decides what data you get.
| Domain property | URL-prefix property | |
|---|---|---|
| Covers | Every subdomain and both http and https | Only URLs starting with the exact prefix |
| Verification | DNS record only | HTML file, HTML tag, Google Analytics, Tag Manager, or DNS |
| Best for | Almost everyone — the complete picture | Reporting on one section, or when you can't touch DNS |
| Watch out | Needs registrar access | www and non-www count as different properties |
Create a Domain property if you can. DNS verification survives a platform migration, a redesign, and a host change — the HTML-file method does not. Many agencies keep both: a Domain property for the true total, and URL-prefix properties for sections they report on separately, like a blog or a store.
Submit your XML sitemap under Indexing › Sitemaps (WordPress and Shopify both generate one — usually /sitemap.xml or /sitemap_index.xml). Then add the people who need access under Settings › Users and permissions. Give clients Full access to their own property rather than sharing a login, and remove contractors when the engagement ends.
New properties start empty. Data begins accumulating from the day you verify and does not backfill, which is the single best argument for verifying a site the day it launches.
DNS record is the most durable: it lives at the registrar, survives a host change, a replatform and a redesign, and is the only route to a Domain property. HTML tag is quickest on WordPress — most SEO plugins have a field for it — but it disappears if the theme or plugin is replaced. HTML file upload works anywhere you can put a file at the root, and breaks the moment someone tidies the server. Google Analytics or Tag Manager verification is convenient if one of those is already installed, and tied to that install continuing to exist.
Whichever you use, leave it in place. Google re-checks verification periodically, and a property that fails re-verification stops collecting data. On a site you will manage for years, spending five minutes with the registrar is the cheapest insurance available.
Shopify exposes a verification field in the theme settings and generates /sitemap.xml automatically. WordPress with Yoast, Rank Math or SEO Framework offers both a verification field and a sitemap index at /sitemap_index.xml. On either platform, confirm the sitemap URL actually loads in a browser before you submit it — a typo here is the most common cause of a "Couldn't fetch" status.
⏱ 4 min read
The Performance report is where you will spend most of your Search Console time. Everything in it is built from four metrics, and the whole report makes sense once those four do.
| Metric | What it counts | What moves it |
|---|---|---|
| Impressions | Times a link to your site appeared in results | Rankings, indexing, demand for the query |
| Clicks | Times someone clicked through to your site | All of the above, plus your title and description |
| CTR | Clicks divided by impressions | Position, snippet wording, competing SERP features |
| Average position | Mean ranking across impressions | Relevance, authority, competition |
Turn on all four metric tiles, set the date range to Compare last 3 months to previous period, then work down: chart first (is the line up, down, or flat?), then the tabs (Queries, Pages, Countries, Devices, Search appearance), then a filter to test whatever the first two passes suggested.
Three filters do most of the useful work. Query contains a question word ("how", "what", "best") surfaces informational demand you could write for. Page equals one URL shows every query a single page ranks for, which is how you find the topic a page has accidentally become known for. Position between 8 and 20 finds pages sitting just off page one — the cheapest wins available to you.
Search Console also breaks out a Search appearance dimension, and since June 2026 Google has been rolling out a separate view for visibility inside AI features. If your property has it, that view is where AI Overviews and AI Mode impressions get isolated instead of blending into the totals.
Performance reports in Search Console — Google Search Central. The official walkthrough of filters, comparisons, and the dimension tabs. Open on YouTube ↗
A single period tells you almost nothing. Search Console's comparison mode is where the report earns its keep: last 28 days against the previous 28 tells you about momentum, and last 28 days against the same period last year strips out seasonality. Most alarming drops turn out to be either a seasonal pattern that repeats every year or the two-to-three-day processing lag at the edge of the range.
When you compare, read the difference column rather than the totals. Sort the Pages tab by click difference and the story usually tells itself in the top and bottom five rows.
The interface caps rows and makes multi-dimensional analysis awkward. Export to Sheets when you want to combine dimensions — query by page, for example — or to keep a monthly snapshot for a client report. For anything ongoing, Looker Studio connects directly to Search Console and rebuilds the same view automatically each month, which is usually worth the half hour it takes to set up once per client.
⏱ 4 min read
Ranking is a later problem. First Google has to crawl the page (fetch it), then index it (store and understand it). A page that is not indexed cannot rank at all, and Search Console gives you two tools to check.
Under Indexing › Pages you get a split: indexed pages and not-indexed pages, with a reason listed for every URL in the second group. The reasons matter more than the count.
| Reason | Means | Do you act? |
|---|---|---|
| Excluded by 'noindex' tag | You told Google to stay out | Only if it was unintentional |
| Alternate page with proper canonical | A duplicate pointing at the original | No — this is working correctly |
| Crawled – currently not indexed | Google looked and passed | Yes — usually a content quality signal |
| Discovered – currently not indexed | Google knows the URL, hasn't fetched it | Yes — often crawl budget or internal linking |
| Blocked by robots.txt | Crawling disallowed | Yes, if it's a page you want found |
| Not found (404) / Soft 404 | Missing, or empty-looking | Yes — restore or redirect |
A healthy site has plenty of non-indexed URLs. Tag archives, paginated pages, thank-you pages and admin URLs have no business in the index. The number to watch is not the total — it is whether pages you care about appear in that list.
Paste any URL from your property into the search bar at the top. You get its index status, the date of the last crawl, the canonical Google chose (which may differ from the one you declared), whether indexing is allowed, and any structured data detected. Test Live URL fetches the page fresh, which is how you confirm a fix before waiting for a recrawl.
If the page is fixed and indexable, use Request Indexing. It is a queue, not a command — it works fine for a handful of pages after a fix, and does nothing useful if you spam it. When a report is running behind, URL Inspection is your fresher source of truth, because it queries the live index rather than the cached report.
Page missing from Google? Inspect the URL. If it says indexed, the problem is ranking, not indexing — go back to the Performance report. If it says not indexed, read the reason, fix the cause (remove the noindex, unblock robots.txt, improve thin content, add internal links), test live, then request indexing.
Occasionally you need something out of Google fast — a staging URL that leaked, a page published early, an old document with a client's details in it. The Removals tool hides a URL from results for about six months. It is a stopgap, not a fix: the underlying page is still indexed. Pair it with the real remedy — delete the page and return 410, add a noindex tag, or put the content behind authentication — or it will reappear when the temporary block expires.
There is no promised timescale. A new page on a site Google crawls daily may appear within hours; a page on a rarely-crawled site can take weeks. Internal links from pages that already get crawled often are the most reliable accelerant, which is another reason to link new work from a hub page rather than leaving it reachable only from a sitemap.
⏱ 4 min read
The left-hand menu holds more than Performance and Pages. Here is what the rest is for, in the order it usually matters.
An XML sitemap is a list of URLs you want crawled. It does not guarantee indexing; it speeds up discovery, which matters most on new sites and large ones. Submit it once. Then check the status occasionally: "Couldn't fetch" usually means a wrong URL or a blocked file, and a discovered-URL count far below your page count means your sitemap plugin is filtering something out.
The Core Web Vitals report groups your URLs by real-world speed data from Chrome users. Three metrics: LCP (main content loaded, aim under 2.5s), INP (responsiveness to clicks and taps, aim under 200ms), and CLS (layout jumping around, aim under 0.1). INP replaced the older FID metric in March 2024 and is the harder one to pass, because it measures every interaction, not just the first.
This report uses field data — what real visitors experienced — which is why it can disagree with a PageSpeed Insights lab score. Field data is the one tied to ranking, and it needs enough traffic to populate.
If Google detects structured data on your pages — FAQs, products, reviews, breadcrumbs, recipes — it adds a report per type showing valid items, warnings, and errors. Warnings are optional fields you skipped. Errors mean the rich result will not appear. Shopify stores get Merchant listings reporting here too.
Two lists: external links (which sites link to you, and which of your pages they point at) and internal links (how your own linking distributes attention). The internal list is the more actionable of the two — an important page near the bottom of it is not getting internal support, and that is free to fix.
Check these when you inherit a site, and otherwise trust the email alerts. A manual action means a human reviewer at Google penalised the site for a guideline violation. It is rare, it is serious, and it is fixed by correcting the issue and submitting a reconsideration request. Empty is the normal state for both reports.
Core Web Vitals reports group similar URLs together, so a single "poor" group can represent hundreds of pages sharing one template. That is useful: fix the template and the whole group moves. It also means the counts jump around as URLs move between groups, which looks like volatility and usually is not.
Field data needs traffic to exist at all. Low-traffic sites will see "insufficient data" for most groups, and there is nothing to fix — use PageSpeed Insights lab diagnostics for direction instead, while remembering the lab score is not the one tied to ranking.
Search Console has added summarised Insights views that blend Search and Analytics data into highlights. They are a reasonable starting point for a client conversation and a poor basis for a decision — the underlying reports are where the detail lives. The same applies to any automated "recommendations" a plugin surfaces: treat them as a list of things to check, not a list of things to do.
⏱ 4 min read
Search Console rewards a short, repeated habit far more than an occasional deep dive. Here is a routine that fits in a coffee break and catches almost everything that matters.
Core Web Vitals trend, Enhancements errors, internal links for your money pages, and — if your property has the generative AI view — how impressions inside AI features are trending against classic results.
Search Console tells you what happened in Google. It cannot tell you what the visitor did afterwards, or whether any of it produced a lead — that is Google Analytics. And it will not tell you why a page ranks where it does; that is the SEO course. Take them in that order and the numbers here start meaning something.
The same fifteen minutes produces a client update if you capture three things each time: what moved, why you think it moved, and what you are doing next. Screenshots of the comparison chart, the top five gaining pages and the top five losing pages carry more meaning than a table of every query. Add a one-line change log — "published the pricing page on the 12th" — and three months later you can attribute results instead of speculating about them.
Google emails property owners about indexing problems, structured data errors, manual actions and security issues. Make sure those go to an address someone monitors rather than a shared inbox nobody opens. For anything more proactive, a scheduled Looker Studio email or a simple monthly calendar reminder beats an unread dashboard.
None of these are difficult to avoid. They are simply the mistakes that come from reading the tool as a scoreboard rather than as a diagnostic, and the routine above is designed to break that habit.
⏱ ~6 min
Answer all ten, then score yourself. Explanations appear after scoring.
All consulted September 2026. Search and AI products change fast — check the primary documentation before acting on anything time-sensitive.