Google Search Console for SEO. The free tool that does most of the SERP-side measurement work.

What Google Search Console actually is

Google Search Console (GSC) is Google's free tool for site owners. It tells you what Google sees on your site, which queries surface your pages in the SERP, how often those pages get clicked, what the average position is, and which technical issues are preventing indexing. It is the closest thing you will get to a direct line from Google about your site, and it is free. There is no paid tier, no usage cap, no enterprise version. Every site should have GSC verified the day it goes live.

The tool covers four domains of information. The SERP-side picture (queries, impressions, clicks, position, CTR by page and by query). The indexing-side picture (which pages are indexed, which are not, and why). The technical-health picture (Core Web Vitals, mobile usability, security issues). The manual-action picture (whether Google has applied a penalty to the site). Each one feeds a different part of the SEO workflow.

What GSC does not cover. It does not show you what happens on your site after the click; that is GA4's job. It does not show you Bing, DuckDuckGo or other search engines; that is what their respective webmaster tools cover. It does not show you keyword-level position over time for specific queries at specific locations; that is what a rank tracker does. The four tools (GSC, GA4, rank tracker, crawler) are complementary, not interchangeable. See the SEO Measurement pillar for the full stack.

Setting it up properly

Most clients arrive with GSC set up in one of three suboptimal ways. URL-prefix property only (missing the non-www variant). One property per subdomain with no consolidated view. The property verified by a team member who has since left the company and nobody else has access. All three are fixable in an afternoon.

Step 1: Verify at the domain property level

Domain properties were added to GSC in 2019 and have been the default ever since. A domain property captures every subdomain (www, blog, shop, app) and both HTTP and HTTPS variants in one consolidated view. URL-prefix properties only capture the exact prefix you verified. For nearly every site, the domain property is the right answer.

To verify a domain property you add a TXT record to the domain's DNS. Most DNS providers (Cloudflare, Route 53, GoDaddy, Crazy Domains) let you do this in under five minutes. Once verified, the domain property captures all subdomains and protocols automatically, and you never have to re-verify when you migrate from HTTP to HTTPS or add a new subdomain.

Step 2: Submit an XML sitemap

In the Sitemaps report, submit the URL of your XML sitemap (usually /sitemap.xml or /sitemap_index.xml). Wait 24 to 48 hours, then check the Sitemaps report again to confirm GSC has processed it and shows the discovered URL count. If the discovered count is wildly different from the URL count in your sitemap, something is wrong with the sitemap. See the XML sitemaps chapter and the sitemap glossary entry.

Step 3: Link GSC to GA4

In the GA4 Admin, under Product Links, link the Search Console property. This surfaces a Search Console reports section inside GA4 (under Acquisition) that combines GSC's query and landing-page data with GA4's session and conversion data. Useful for spotting which queries are driving converting traffic rather than just clicks.

Step 4: Tag your brand terms

Document the set of branded queries up front. Company name, common misspellings, product names, founder name if it gets searched. You will use this list to build a regex filter that separates branded from non-branded queries on every Performance report. We do this on day one of every engagement and revisit the list quarterly. See the branded versus non-branded section below for the exact regex pattern.

Step 5: Set up the BigQuery export

GSC's interface only shows 16 months of historical data. After that, the data is gone. The fix is the bulk-data export to BigQuery, which Google added in 2023. Set it up on day one of a new property and you will have unlimited history for year-on-year comparisons three years from now. The export is free up to BigQuery's free tier, which covers nearly every SMB.

Step 6: Subscribe to indexing and manual-action alerts

GSC sends email alerts for various issues. Most are noise (a single mobile usability warning, a Core Web Vitals threshold drift). Two categories are worth catching immediately: indexing issues that affect a meaningful share of pages, and manual actions. Both can crater traffic. Everything else can wait for the monthly review.

The four reports that matter

GSC has roughly twenty reports if you count the sub-views. Four of them do most of the actual work. The rest are diagnostic.

Report 1: Performance (Search results)

The single most-used report in GSC. Shows clicks, impressions, average position and click-through rate (CTR) for every query and every page over the chosen time range. Default view is the last 3 months; we typically run a 28-day rolling versus the prior 28-day rolling comparison for trend analysis.

The Performance report has four tabs underneath the chart: Queries, Pages, Countries, Devices. Each one slices the same underlying data by a different dimension. The Queries tab is where most of the analysis happens; the Pages tab is where you find the landing-page-level winners and losers. Use the date range comparison feature to spot week-on-week and month-on-month movement.

The most common mistake on the Performance report is reading the headline numbers without applying any filter. Total clicks across all queries blends informational, commercial and branded traffic into a single number that means very little. Always filter by query intent (branded versus non-branded), by page type (commercial landing pages versus blog posts), or by country (for sites with international traffic). See the SERP performance chapter for deeper analysis patterns.

Report 2: Page Indexing

Shows which of your pages are indexed, which are not, and why. The "Why pages are not indexed" section is the most useful bit; it groups un-indexed pages by reason (excluded by noindex tag, duplicate without user-selected canonical, crawled but not indexed, soft 404, server error, redirect). Each group is a separate troubleshooting workflow.

Three patterns worth watching. A sudden spike in "Crawled - currently not indexed" usually signals a quality or duplication issue. A spike in "Discovered - currently not indexed" usually signals a crawl budget or internal-linking issue. A spike in "Soft 404" usually signals a thin-content or template problem. The technical audit chapter covers the workflow for diagnosing each one. See also crawling, indexing and ranking for the foundational concepts.

Report 3: Core Web Vitals

Shows the field-data performance of your site across LCP, INP and CLS. Pages are bucketed as Good, Needs Improvement or Poor based on the 75th-percentile real-user data Google collects via Chrome. Mobile and desktop are reported separately.

The trap on this report is reading it as if it were lab data. Field data lags by 28 days, so changes you made yesterday will not show up for a month. The other trap is over-reacting to a few pages slipping into Needs Improvement; the question that matters is whether your highest-traffic pages are in Good. See the Core Web Vitals glossary entry and the dedicated Core Web Vitals pillar.

Report 4: Manual Actions and Security Issues

Should be empty. If it is not, the site has been penalised by a Google review or has a security compromise. Manual actions usually relate to unnatural links, thin or scraped content, hidden text, or structured-data abuse. Security issues are usually malware or phishing flags.

Both are emergencies. Manual actions tank traffic until cleared. Security flags drop the site out of the SERP entirely until resolved. If the report shows anything other than "No issues detected", stop what you are doing and triage immediately. See the penalty glossary entry and Off-Page SEO pillar for the recovery workflow.

Branded versus non-branded

Most SEO reports we inherit ignore the branded-versus-non-branded split. The unsegmented "total organic clicks" number then drives the conversation. The number lies, because brand demand and SEO progress are different things.

Brand demand mostly comes from offline marketing, PR, paid media, word of mouth, and existing customers. SEO work mostly grows non-branded organic clicks (people searching for the service or topic, not for your specific company name). The two move on different cycles and respond to different inputs. Lumping them together hides the picture.

The fix is a regex query filter applied to every Performance report you run. In the Performance report's query filter, choose "Custom (regex)" and use a pattern like the one below. The exact list depends on the client.

For a fictional Perth plumber called "ABC Plumbing", with founder name Joe Bloggs and a few common misspellings, the regex would catch any query containing "abc plumbing", "abcplumbing", "abc plumb", "abc plumber" or "joe bloggs". Apply that pattern with the "Doesn't match" option to surface non-branded queries; flip to "Matches" to surface branded queries. Two separate reports, two different stories.

The typical pattern across Perth and WA clients: branded queries are usually 30 to 60 percent of total clicks for established businesses. New businesses have almost no branded volume. Service businesses with a recognisable name have more branded; commodity-product e-commerce has less. The split itself is interesting; the trends within each segment are more interesting still.

The honest SEO progress conversation tracks non-branded clicks, non-branded conversions, and the diversity of non-branded queries driving traffic. Branded growth is good news but rarely attributable to SEO work; non-branded growth is the SEO scorecard. See SEO KPIs for the wider KPI structure and search intent for how to read query intent properly.

The limits of GSC data

GSC is the single most valuable free SEO tool, but it has six limits worth understanding.

• 16 months of history. The interface caps at 16 months. The BigQuery export removes this limit if you set it up early.
• Query data is sampled and anonymised. GSC suppresses some queries to protect user privacy, particularly long-tail queries with low volume. Total clicks are accurate; per-query clicks for the long tail are under-reported.
• Average position is averaged. A query that ranks position 1 sometimes and position 50 other times shows up with an average position around 25. The number is honest but misleading without a distribution view.
• No conversion data. GSC shows clicks; it does not know what happened next. You need GA4 or your CRM to close the loop.
• SERP feature impressions are weird. Featured snippets, sitelinks, AI Overview citations and other features each have slightly different impression-counting rules. The result is that a SERP-feature-rich query can show inflated impressions versus the underlying ranking.
• Discovery delay on new content. New pages take a few days to a few weeks to start showing in the Performance report. Do not interpret "no data yet" as "no traffic".

None of these are dealbreakers. They are just shape-of-the-tool things to remember when reading the data.

A monthly GSC workflow

The same five-step workflow we run on every client GSC account on the first business day of the month.

1. Performance report, 28-day rolling versus prior 28-day rolling. Record total clicks, total impressions, average position, average CTR. Split by branded versus non-branded using the regex filter. Note the top movers (positive and negative) at the query level and the page level.
2. Page Indexing report. Record the indexed page count and any change. Check the un-indexed categories for new issues. Triage anything that has grown significantly.
3. Core Web Vitals. Check whether the share of Good pages has held steady or shifted. Investigate any pages that have slipped from Good to Needs Improvement.
4. Manual Actions and Security Issues. Confirm both are clean. If they are not, drop everything and triage.
5. Quick scan of any new alerts since the last review. Most will be noise. Anything related to indexing or manual actions gets investigated immediately.

The whole workflow takes 30 to 60 minutes for a typical SMB property. Bigger sites take longer. The discipline is doing it on the same day every month so the comparisons are clean.

Common mistakes

Perth and WA context

Two GSC patterns specific to Perth and WA businesses.

Local-intent queries are heavily location-suffixed. "Plumber Fremantle", "electrician Joondalup", "tree lopping Mandurah". These suffixed queries drive most of the commercial volume for local service businesses, but in GSC they get split across dozens of suburb variants. Use the Queries tab's regex filter to group them ("plumber.*(fremantle|joondalup|mandurah)") for a clearer view. The aggregate picture across suburb-suffixed queries is usually much bigger than the per-suburb numbers suggest. See Local SEO and Local SEO Perth for the wider local-search context.

Industry-specific verticals show large query-suppression gaps. Niche industries (mining, marine, agriculture) have lots of long-tail technical queries that get suppressed in GSC for privacy. The visible Performance report under-counts the long tail. We supplement with rank-tracker keyword research and Semrush organic keyword reports to fill the gap. See mining SEO and the Keyword Research pillar.

For the wider context, the GA4 for SEO chapter covers the on-site analytics layer that pairs with GSC, the SEO KPIs chapter covers how to turn GSC data into a defensible KPI structure, the SERP performance chapter goes deeper on SERP analysis, the rank tracking chapter covers how third-party trackers complement GSC, and the title tags chapter covers the on-page lever that most directly affects GSC CTR. For sites worried about AI Overview impact on GSC clicks, the AI Overviews chapter covers the click-curve change. New properties should start with a free SEO audit, and clients ready for a full review can engage the website audit service.