Google Search Console: How to Use It and What to Check Every Week

Google Search Console: How to Use It and What to Check Every Week

Google Search Console is a free Google service that shows how the search engine sees your website: which queries bring people to it, which pages made it into the index and which did not, and what exactly is getting in the way. To get real value from it you need three things: verify ownership, open two reports once a week (Performance and Page indexing), and read the numbers correctly, because there are a few traps. Below we walk through every section of the interface in business-owner language and finish with a six-point weekly checklist.

Marketer reviewing a Google Search Console report on a laptop in an office

Why you need Search Console when you already have Google Analytics

Analytics counts what happened after the click: how many people arrived, what they did, whether they bought. Search Console shows what happened before the click: how many times your site appeared in results, at which position and for which query. It is the only source of search query data that comes straight from Google rather than from a third-party estimate. It is also where the messages about manual penalties, security problems and mass indexing errors land. The service is free, and signing up at search.google.com/search-console takes a few minutes.

How to verify site ownership

Google has to make sure the site is yours. The method depends on what you have access to. At Query we recommend adding a Domain property through DNS: it covers every subdomain, http and https at once, and it survives a redesign.

MethodWhat you needWhen it fits
DNS recordTXT record at the registrarMain option: all subdomains and protocols
HTML fileFile in the site rootFile access, no DNS access
HTML tagmeta in the homepage headCMS with a header code field
Google Analytics"Edit" rights in GAAnalytics already installed
Google Tag Manager"Publish" rights in GTMTags managed via GTM

With the file and tag options, verification only covers one specific prefix, say https://example.com/. If part of your traffic lands on www or on http, those addresses will not show up in the reports. After verification the data starts from zero, there is no history from before you connected, so it pays to connect on day one of a site's life.

The Performance report: how to read clicks, impressions, CTR and position

This is the main report and the one you open first. The four metrics at the top mean the following:

  • Impressions: how many times a link to your site appeared in search results. The person may not even have seen it if it sat at the bottom of the page.
  • Clicks: how many times it was clicked.
  • CTR: the share of impressions that turned into clicks. Normal values for position one are several times higher than for position ten.
  • Average position: the site's averaged rank in results over the selected period and set of queries.

Filtering by queries and pages

The power of the report is in its filters. The Queries tab answers "which words do people use to find us", the Pages tab answers "which pages earn the traffic". Click a specific page in the list and the query table rebuilds for that page alone, which shows exactly what a product page or an article ranks for. A "Query contains" filter with your brand name separates branded traffic from generic search, and the Compare toggle puts two periods side by side. The most useful combination for a weekly review is a page filter plus a comparison of the last 28 days against the previous 28.

Clicks and impressions chart in the Google Search Console Performance report on a monitor

The report splits every address Google knows about into two groups: indexed and not indexed. The mere fact that some pages are out of the index is not a problem: pagination, filter pages and service URLs should not be there anyway. The problem is when pages that make money end up in the "not indexed" list. The most common reasons and what they mean:

  • Crawled, currently not indexed: Google saw the page but does not consider it valuable enough. Usually thin or duplicated content.
  • Discovered, currently not indexed: Google knows the URL exists but has not got to it yet. On large sites this is a crawl budget question; on small sites it is a sign the page has no internal links pointing at it.
  • Page with redirect and Alternate page with proper canonical tag: normal, as long as the redirects and canonicals were set deliberately.
  • Excluded by noindex tag: check whether the tag was left over from development. We regularly see online stores where categories have been "temporarily" closed to indexing for six months.
  • Not found (404) and Server error (5xx): sound the alarm if these are pages with traffic.

Clicking a reason opens the list of specific URLs, and the Validate fix button asks Google to re-check them after you make changes. A large share of these errors gets closed during technical SEO optimization, when the site is cleaned of duplicates and the indexing rules are set properly.

Leave an application

Enter your name and email, our managers will contact you as soon as possible.
By submitting the form you agree to our privacy policy.

Core Web Vitals

The Experience section shows whether the site meets three speed and stability metrics based on real visitor data: LCP, INP and CLS. What matters here is not the raw numbers but the groups of "poor" mobile URLs: if a category or product page template falls into one, hundreds of pages suffer at once. We explained what the metrics mean and how to fix them in a separate article on Core Web Vitals.

Links

This report shows which sites link to yours, to which pages and with what anchor text. The list is incomplete compared with paid tools, but it comes straight from Google. Checking it once a month is enough: a sudden jump in linking domains from odd sites deserves attention, and your most-linked pages hint at which content is worth developing.

URL inspection and sitemaps

The search bar at the top of the interface accepts any address on your site and tells you whether it is indexed, when Google last crawled it, which canonical version it chose and whether the page passes the mobile usability check. The Test live URL button checks the page right now, and Request indexing puts it in the queue. This works for new or updated pages, but it does not perform miracles: if Google has already rejected a page as low value, asking again changes nothing.

Submit your sitemap.xml in the Sitemaps section. Once added, Search Console shows how many URLs it read and whether there were errors. The gap between the number of addresses in the sitemap and the number of indexed pages is a useful metric in its own right: if the map lists 5,000 URLs and the index holds 900, something is wrong with the site.

What to check every week: a six-point checklist

One session of 15-20 minutes. Doing it more often makes little sense, because the data updates with a delay and day after day you will be looking at the same chart.

StepWhere to lookSignal
1. Clicks and impressions, 7 daysPerformance, compareDrop beyond seasonal swings
2. New indexing errorsPage indexing404 or 5xx on pages with traffic
3. High impressions, low CTRPerformance, QueriesPosition 3-10, CTR below neighbours
4. Lost positionsPerformance, PagesMinus 3 or more positions
5. Messages and manual actionsBell, SecurityAny new notice
6. Core Web VitalsExperienceURL group moved to "needs improvement"
  • Step 3: rewrite title and description on pages with impressions but few clicks.
  • Step 5: hacks, penalties and sitemap problems arrive here.
  • Step 6: check after every site update.
Team discussing a weekly Search Console report with click and position figures on a screen

Common mistakes when reading the data

Average position is not a position. If a site ranks 2nd for one query and 40th for another, Console shows 21. A single page can have position 1.3 in Ukraine and 45 worldwide. Look at position only for a specific query and country, otherwise the number means nothing.

A 2-3 day lag. Today's figures in Performance are incomplete, so the last few days on the chart always look like a slump. Compare periods that ended at least three days ago.

Impressions without clicks are not always bad. A query like "instagram login" that your blog accidentally ranks for will produce a million impressions and zero clicks. CTR in such cases says nothing about page quality.

The query total is smaller than the overall total. Console hides part of the rare and anonymized queries, so the query table covers only a portion of the clicks. That is normal.

Only 16 months of data. There is no longer history in the interface. If you need year-over-year trends, export the reports or set up the BigQuery export in advance.

The wrong property. Accounts often hold two properties: https://example.com/ and the domain. Their figures differ because the first one sees a single prefix only. Work with the Domain property.

Where to start

  1. Verify the site through DNS as a Domain property and submit the sitemap.
  2. Open Page indexing, go through the reasons and check whether any category or product pages sit among the non-indexed URLs.
  3. In Performance, find the 10 pages with the most impressions and record their CTR and position as a baseline.
  4. Block one hour a week in your calendar for the checklist above.
  5. If the indexing errors are many and it is unclear where to begin, that is a job for an SEO team: at Query this kind of review is part of the first month on any project.

Other articles

Other services

Get in touch
Messengers