Skip to content
Doctor SEO
Free SEO audit

Google Search Console: Verify Your Site and Use It Properly From Day One

How to set up Google Search Console as a Domain property, the settings that matter, five SEO workflows that end in a decision, and what the tool cannot tell you. Verified 10 October 2026.

Google Search Console is the only place where Google itself tells you how your site performs in its search results: which queries showed your pages, which URLs it indexed and why it skipped the others, and whether a human reviewer has taken a manual action against you. It is for anyone who runs a website, from a one-person blog to an in-house SEO team. Google describes it as “a free service offered by Google”, and that was still the case when checked on 10 October 2026: there is no paid tier.

The verdict first: set it up on day one, as a Domain property verified by DNS, and turn on the BigQuery bulk export the same week. Search Console keeps only 16 months of data, and the bulk export does not backfill. Your own uncapped archive starts on the day you switch it on; everything before that exists only in the capped reports and API, and only for 16 months.

How to get access: properties and verification, step by step

Search Console works on properties, and the property type you pick decides what you see. Google’s help page on adding a website property defines two types for websites. A Domain property “includes all subdomains (m, www, and so on) and multiple protocols (http, https, ftp)”. A URL-prefix property “includes only URLs with the specified prefix, including the protocol”, so https://example.com/ does not cover http:// or https://www.example.com/.

  1. Open Search Console, open the property selector dropdown and choose + Add property.
  2. Choose Domain and type the root domain only, without protocol, path or “www”.
  3. Copy the TXT record Search Console shows you and add it at your DNS provider.
  4. Return and click Verify, or use Verify later to save progress. Google warns that manually added records can take “up to two or three days” before your provider serves them.
  5. Keep the record. Google’s verification page says: “To stay verified, don’t remove the DNS record from your provider, even after verification succeeds.”

One documented detail most setup guides skip: data collection for a property starts when anyone adds the property, even before verification. So adding the property today and finishing the DNS step tomorrow loses nothing.

Which verification method works for which property

Google’s verification help page states that DNS is “the only way to verify a Domain property”. Every other method works on URL-prefix properties only.

Method Domain property URL-prefix property What Google requires
DNS record (domain name provider) Yes, the only option Yes Access to your DNS; keep the record after verifying
HTML file upload No Yes The file stays on the server; Search Console does not follow redirects to another domain to find it
HTML meta tag No Yes The tag sits inside the <head> of the homepage
Google Analytics No Yes “Edit” rights on the Analytics account, same Google account in both tools, snippet unmodified
Google Tag Manager No Yes Publish or Admin permission on the container

Verification is not a one-off event. Google says Search Console “periodically checks if your verification token is still present and valid”, and that if the token disappears, permissions expire after a grace period.

Google’s property help page also describes a newer type it is rolling out gradually, platform properties: “your presence on supported platforms like Instagram, TikTok, X, and YouTube”.

The settings that matter, and the ones to leave alone

Six settings change what you can do a year from now. The rest of the Settings page can stay on its defaults.

Setting (where) What to do Documented reason
Property type (+ Add property) Domain property as the main view Covers every subdomain and protocol in one place
Users and permissions (Settings > Users and permissions) At least two verified owners; give agencies and freelancers Full user, not Owner If all verified owners are removed, “all remaining users and delegated owners will lose access to the property after a grace period”
Sitemaps report Submit every XML sitemap or the sitemap index Enables the per-sitemap filter in the Page indexing report
Bulk data export (Settings > Bulk data export) Set it up in week one No backfill; the first export runs up to 48 hours after setup
Associations (Settings) Link the GA4 property Lets Analytics show Search Console data next to behaviour data
Email notifications Make sure the owners’ inboxes receive them Google says there is “no need to sign in to the tool every day” because it emails you about new issues

The permission levels are worth knowing precisely. Google’s users and permissions page says a Full user “has view rights to all data and can take some actions”, such as submitting sitemaps and using URL Inspection, while a Restricted user “has simple view rights on most data”. A property can have a maximum of 100 non-owners.

What to leave alone. The Removals tool gives what Google calls “only a temporary removal of about six months”, so it is an emergency brake, not a cleanup tool. Change of address is only for moving between domains or subdomains, and Google explicitly says not to use it for an http-to-https move. Deleting a sitemap from the report does not make Google forget it either: Google states that “Google won’t forget the sitemap or any URLs listed in it”.

Bulk data export: the setting with a deadline

The bulk export sends performance data every day to a BigQuery dataset you own. Google’s setup page requires a Google Cloud project with billing, the BigQuery API and BigQuery Storage API enabled, and the service account [email protected] added with the BigQuery Job User and BigQuery Data Editor roles. The dataset name must start with searchconsole, and Google warns that “you can’t easily change this location later once your exports have begun”.

Google Cloud storage and query costs apply, with “a free usage level”. Tables are “retained forever, by default”, and Google recommends expiring partitions rather than whole tables, because a table expiry “will delete all your data”.

Five SEO workflows that end in a decision

Each workflow uses only Search Console and ends in an if-then decision.

1. Triage “Crawled – currently not indexed”

The Page indexing report lists this reason when, in Google’s words, “the page was crawled by Google but not indexed”, and it “may or may not be indexed in the future”. Google adds that there is “no need to resubmit this URL for crawling”, and states the broader principle: “Don’t expect every URL on your site to be indexed.”

  1. Open Indexing > Pages, click the reason, and export the example URLs.
  2. Sort them by template: product, category, tag, filter, article.
  3. Inspect two or three URLs per template with URL Inspection.
  4. If the URLs are filters, tags, internal search or near-duplicates, then treat the status as correct and stop linking or listing them in sitemaps.
  5. If they are pages you want ranking (core products, key articles), then the problem is the page, not the crawler: improve or consolidate the content and strengthen internal links before asking Google again.
  6. If a whole template appears here, then run workflow 5 on it before touching individual URLs.

2. Find the query/page pairs with a CTR gap

The Performance report defines CTR as “the click count divided by the impression count”. The useful comparison is your own pages against each other, not an industry benchmark, because no published CTR curve applies to your queries.

  1. In Performance > Search results, set the search type to Web: text-based (multimodal traffic has no query data), choose the last three months and turn on all four metrics.
  2. Filter one important page, open the Queries tab and sort by impressions.
  3. Note queries where average position sits on page one but CTR is clearly below what the same page earns on its other page-one queries.
  4. If position is on page one and CTR lags, then the snippet is the suspect: rewrite the title and meta description, record the date, and compare the next 28 days against the previous 28 with the date comparison.
  5. If position is beyond page one, then the snippet is not the problem, and rewriting it will not fix ranking.

Keep in mind that filtering by query or page affects totals, because, as Google’s filtering page puts it, of “data truncation and the omission of anonymized queries”.

3. Compare sitemap URLs with indexed URLs

Your sitemap is your list of pages you want indexed, and Google’s Sitemaps report says plainly that “there is no guarantee that a page URL discovered in a sitemap has been or will be crawled or indexed”. The Page indexing report has a filter for exactly this check.

  1. In Indexing > Pages, switch the filter from “All known pages” to “All submitted pages”.
  2. Compare indexed against not indexed for the submitted set, then switch to “Unsubmitted pages only”.
  3. If many submitted URLs are not indexed, then read the reasons: redirects, noindex or canonicalised URLs in a sitemap mean the sitemap is lying to Google, and it should list only canonical, indexable 200 URLs.
  4. If many unsubmitted URLs are indexed, then either they belong in the sitemap or they should not exist as separate indexable URLs.

4. Check a site after a migration

Google’s site move guide (Last updated 2026-08-20 UTC) tells you to submit both the old and the new sitemaps and to expect indexed counts to fall on the old site and rise on the new one. For medium-sized sites, “it can take a few weeks or more for Google to gradually start showing the new URLs”.

  1. For a domain change, run the Change of Address tool from the old domain-level property. You must own both properties with the same Google account.
  2. Submit the old and new sitemaps in the new property.
  3. Each week, compare Page indexing for both properties and Performance by page on the new one.
  4. If old URLs drop out but new ones do not climb after several weeks, then inspect a sample of old URLs and confirm each returns a single 301 to its equivalent, not to the homepage.
  5. If you are deciding when to remove redirects, then keep them for at least a year.

That last step rests on a documented discrepancy worth knowing. Google’s Change of Address help page says to keep redirects “for at least 180 days”, while the site move guide says to keep them “for as long as possible”, which is “generally at least 1 year”. Both are Google documents checked on 10 October 2026; following the longer one satisfies both.

5. Run URL Inspection on a key template

URL Inspection shows what Google’s index holds for one URL: last crawl, the user agent used, the user-declared canonical and the Google-selected canonical. Google’s URL Inspection page warns that the indexed view “is not a live test”, and that the live test does not check “whether this is a duplicate or alternate page”, because canonical choice “is determined at indexing time”.

  1. Pick one representative URL per template: home, category, product or article, and paginated listing.
  2. Inspect it and compare User-declared canonical with Google-selected canonical.
  3. Click Test live URL, then View tested page, and read the screenshot and the rendered HTML.
  4. If the canonicals differ, then the template sends conflicting signals (internal links, redirects, sitemaps or duplicated content) and needs fixing at template level.
  5. If the rendered HTML lacks main content that a browser shows, then the content depends on rendering that Google did not complete in the test, and it should be checked with your developers.

Request indexing has a daily limit and “does not guarantee that the page will appear in the Google Index”: use it for a page you just fixed, not as routine.

The day-one checklist

Run this once per site; each item protects data or access you cannot recover later.

  1. Domain property added and verified by DNS TXT record; record left in place.
  2. A second verified owner on the property.
  3. Agency or freelancer access set to Full user.
  4. XML sitemap or sitemap index submitted, status “Success”.
  5. Bulk data export to BigQuery configured, partition expiry set rather than table expiry.
  6. GA4 property linked under Associations.
  7. Manual actions and Security issues reports opened once, so you know what “no issues” looks like.
  8. One URL per template inspected, canonicals compared.

Reports you will open less often

The Core Web Vitals report uses field data from the Chrome UX Report (CrUX) and groups URLs with a similar experience. Google’s report page sets the “good” thresholds at LCP ≤2.5s, INP ≤200ms and CLS ≤0.1, and a URL group’s status is “its most poorly performing metric”. How much these metrics weigh in ranking is disputed: one camp treats them as a major factor and sells audits on that basis, another holds that the ranking effect is small and the real case is user experience and conversion. No independent study has isolated the weight, and Google’s report page makes no statement about ranking at all.

The Manual actions report shows when “a human reviewer at Google has determined” that pages violate Google’s spam policies, and Google’s page warns that “fixing the issue on just some pages will not earn you a partial return to search results”. The Security issues report covers hacked content, malware and unwanted software, and social engineering. Both end in Request Review, and Google says reviews “can take several days or weeks”.

The Crawl Stats report (Settings > Crawl stats) shows total crawl requests, download size, average response time and host status for robots.txt fetching, DNS and server connectivity. Google says it is “available only for root-level properties”, so it needs a Domain property or a root-level URL-prefix property. The Insights report gives a simplified overview with top and trending content and queries, and Google notes that trending order is based on the number of clicks, not the percentage change.

Limits: what Search Console cannot tell you

Search Console is Google’s own data, but sampled, truncated and capped, as the help pages admit.

The September 2026 multimodal split: an open question

On 24 September 2026 Google split the Web search type into Web: text-based and Web: multimodal. The Performance report help page now says multimodal “includes search traffic when users search using images (such as with a smartphone camera)”, covering Google Lens, Circle to Search, uploaded images and Chrome image search, and that “the queries dimension isn’t available when this search type is selected”.

Whether those rows are new data or data moved out of the old Web total is disputed, and the published evidence does not settle it. Google’s position, given by John Mueller on Bluesky and reported by Search Engine Roundtable, is that “the data wasn’t previously in the counts, the report has new data.” Against that, Dave Smart told the same outlet: “I am seeing corresponding drops in impressions and clicks starting the 10th on a few properties.” He disclosed no property count or volumes. PPC Land notes that the cause of those drops remains unidentified. Nobody has published a paired impression series across both dates with the properties disclosed, so the data cannot settle it yet. If you want to look at this on your own site, one way is to compare your export across 10 and 24 September; neither camp has published that comparison.

One fact is not in dispute. When checked on 10 October 2026, the Search Analytics API reference (Last updated 2026-08-11 UTC) listed the type values discover, googleNews, news, image, video and web, and the word “multimodal” did not appear. The split exists in the interface, not in the API, so dashboards built on the API cannot reproduce it.

Alternatives and complements

Nothing substitutes for Google’s own data, so the alternatives are complements.

Question Search Console Complement
What happens after the click? No Google Analytics 4
How does Bing see the site? No Bing Webmaster Tools (our tool page publishes on 29 October)
Every URL and link on the site, crawled now Samples only A desktop crawler such as Screaming Frog
Competitors’ rankings and links No Third-party SEO suites, which model these figures from their own data

Every tool this section covers is on the SEO tools hub.

Where the course goes deeper

This page owns the tool. The method, meaning how to read search data without fooling yourself, belongs to the course. The SEO analytics level covers measurement, the technical SEO level covers crawling, indexing and canonicalisation and the lesson on connecting Claude to Search Console and GA4 via MCP shows how to query your own property in plain language.

Sources