The typical interaction with Search Console: open it, look at the clicks graph, sigh, close it. Meanwhile it holds the only honest record of your search traffic — the pages that lost half their clicks last quarter, the queries sitting at position 11 one step outside the first page. Below: how to connect it without the usual mistakes, which cuts to open monthly, and what you can safely ignore.

Connecting it without the mistakes that hurt later

The first decision is property type, and it is the one most often got wrong. A Domain property collects data across every subdomain and protocol; a URL-prefix property covers only the exact address. Add https://site.com/ instead of the domain and you will never see traffic from www or subdomains, while sincerely believing there is none.

1. 1

   Add a Domain property, not a prefix

   Verification is a TXT record in DNS — five minutes in your registrar panel, and it settles www and https duplicates immediately. Keep URL-prefix properties for when you want one section reported separately.

2. 2

   Submit the sitemap

   Sitemaps section, full sitemap.xml URL. Check a day later how many URLs were read: if the file lists 400 pages and 120 were discovered, the sitemap is the problem, not indexation.

3. 3

   Link it to GA4

   Do this from the analytics side, not from Search Console. Once linked, a query report appears in GA4, so you see not just the click but what happened next.

4. 4

   Hand out access correctly

   Give contractors a Restricted or Full user role. They do not need owner rights: an owner can delete the property and add their own verifications.

The Performance report: what those four numbers really mean

The graph with four toggles is what everyone looks at and almost everyone misreads. The data is fine; the trouble is that three of the four metrics behave differently from how they appear.

Second thing worth learning early: the totals in the query table never match the headline number. That is not a bug — rare queries are anonymised and never appear as rows, though they still count towards the total. Work with relative change rather than balancing the books.

The most useful control in the report is the regex filter. Without it you are staring at a blend of branded traffic, which would have arrived anyway, and genuine search demand.

# Performance → Query → Custom (regex)

# 1. Non-branded queries: exclude the company name and variants
Doesn't match regex:
(?i)webengine|web engine|web-engine

# 2. Question queries — candidates for an FAQ block and AI answers
Matches regex:
(?i)^(how|why|what|when|which|can|should)\b

# 3. Commercial intent
Matches regex:
(?i)(price|cost|hire|buy|quote|agency)

Five data cuts that find lost clicks

This is the part worth opening every month. Each item is a specific sequence of filters and a specific action, not an instruction to "analyse the data".

1. 1Queries at positions 5–15 with heavy impressions. Filter by position, sort by impressions. These pages are one step from where clicks begin: reworking the title and opening paragraph pays off in two to four weeks.
2. 2Abnormally low CTR at a good position. A page sitting third or fourth with CTR under 3% has a snippet problem, not a ranking problem. Rewrite the title and description around query intent, re-check in a month.
3. 3Last 3 months against the previous 3. Pages tab, comparison mode, sorted by click difference. Anything down more than 30% is a candidate for a refresh, not for writing something new.
4. 4One page serving two intents. If the URL report mixes informational and commercial queries, you have cannibalisation. Split it in two — total traffic usually rises even though each page looks weaker alone.
5. 5Queries with no matching page. Filter by topic and check the Pages column: the homepage or a random section means demand exists with nothing built for it. That is an instruction to create a page, backed by your own data.

Indexation: the Pages report and URL inspection

The second most important section, and the first to open after every release. The status labels are not self-explanatory, so here is the plain-English translation.

URL inspection shows how Google sees one page: when it was crawled, which canonical it chose, whether structured data validated. "Request indexing" is useful after real changes, but the quota is a handful of URLs per day, so it cannot bulk-solve indexation. If dozens of pages are stuck, that is grounds for a technical audit rather than manual submissions.

A monthly routine, and what you can ignore

Watching the reports constantly is harmful: search traffic swings 15–20% week to week for no reason, and reacting to that wastes time. The workable rhythm is once a month along a fixed route, plus a check after every release.

One boundary worth remembering: Search Console knows nothing about what happens after the click. How long the visitor stayed, whether they reached the form, whether they came back — that is analytics territory, and without it the picture is half-finished. Joining both sources is covered in our analytics setup guide.

Frequently asked questions about Google Search Console

In short

• Add a Domain property via DNS — a URL-prefix property hides part of your traffic.
• Average position is the least useful metric; work with clicks, impressions and CTR by query group.
• The fastest wins live at positions 5–15, where a reworked page shows results in two to four weeks.
• Check the Pages report after every release — accidental noindex tags surface there first.
• Review monthly along a fixed route; 15–20% weekly swings mean nothing. The rest is covered by the general SEO checklist.