GA4 vs Search Console: Why Your Numbers Never Match (and Which One to Trust)

Short answer: GA4 and Search Console never match because they count different things at different moments. Search Console counts a click on Google's results page. GA4 counts a session only after your page loads and its tag runs. Google calls Search Console the source of truth for search performance and Analytics the source of truth for what happens on your site.

Over the same 90 days this autumn, our own Search Console reported 719 clicks. Our GA4 reported 1,349 Organic Search sessions. That is nearly double, and it runs the wrong way. Most articles on this topic tell you that GA4 will show less than Search Console, because of ad blockers and cookie banners. We run a cookie banner that denies analytics until a visitor accepts, so ours should have come in low.

We are telling you this because the textbook explanation is only half the story. The two tools can disagree in either direction, and the direction tells you something. Here is how to read it.

Two witnesses standing in different rooms

Think of a shop with a doorman on the street and a cashier at the till. The doorman counts everyone who turns in from the road. The cashier counts people once they reach the counter. Some people turn around in the doorway. Others come in through the side door from the car park, which the doorman never sees. Nobody would expect their totals to match, and nobody would call either of them a liar.

Search Console is the doorman, and he only watches one street: Google. GA4 is the cashier, and she sees everyone who reaches the counter from any direction, but only if her till is switched on. Google's own page on using the two tools together makes the same split and names clicks and sessions as the closest comparable pair.

Why GA4 usually comes in lower

Google lists several reasons for large gaps. These are the ones that push GA4 down.

• The tag did not run. A visitor clicks, then leaves before the page loads, or a page is missing the tag. Search Console has already counted the click.
• Consent. If someone declines analytics cookies, GA4 does not record them in the normal way. Behavioural modelling can estimate those visitors, but only on properties with at least 1,000 events a day from people who declined, for at least 7 days, and at least 1,000 daily users who accepted on 7 of the previous 28 days. A small business site rarely qualifies.
• Files that are not web pages. A click to a PDF counts in Search Console, but the file has no tag to fire.
• Bots. GA4 leaves out known bots and spiders.

Why GA4 can come in higher, as ours did

Our 1,349 sessions against 719 clicks needs its own explanation. We have not proved the exact share of each cause, so read these as likely contributors, and note that the two date windows differ by a couple of days.

• Other search engines. GA4's Organic Search channel includes any engine on its list, including Bing and Yahoo. Search Console reports Google alone.
• Sessions restart. A GA4 session ends after 30 minutes of inactivity. If someone clicks once and comes back an hour later, that is one click and potentially two sessions. Our 1,349 sessions came from only 574 users, or about 2.4 sessions each, which is a strong hint.
• Attribution. GA4 credits sessions using non-direct last click, so a returning visitor who types your address can still be credited to the earlier organic source. Search Console simply counts every click.

That sessions-per-user ratio is the sort of detail worth checking on your own site. It takes ten seconds and explains more than any general list.

How to compare them properly

If you want a fair comparison, take away the differences you can control.

1. In GA4, open Traffic acquisition and filter to Session source "google" and Session medium "organic".
2. Use the same dates in both tools, ending three days ago. Search Console data is usually available in 2 to 3 days.
3. Compare by landing page, not just the site total. Pick your top five pages.
4. Compare direction across months, because the trend matters more than any single figure.

The easiest way to do this inside GA4 is to link Search Console to your GA4 property. The linked reports do have limits. Only landing page, device and country work with Search Console metrics, and the data arrives 48 hours after collection.

When a gap is a warning

A steady gap is boring, and boring is good. Stop being relaxed when the gap changes suddenly. Four patterns deserve attention:

• Search Console flat, GA4 falling. Your tag broke or your consent setup changed. This is the classic signal, and our traffic drop check starts here.
• Direct rising as Organic falls. Sessions are losing their source. This can follow a redirect that strips referrer data. In the last 90 days, 57% of our GA4 sessions were Direct, which is high enough that we treat that bucket with suspicion.
• Unassigned growing. When session source shows as (not set), the traffic lands in Unassigned. Missing session_start events are a common reason.
• One page out of line. Check whether the tag is on that page and whether Google has chosen a different canonical URL.

Which tool answers which question

Most arguments about these numbers happen because someone asks one tool a question that belongs to the other. Use this as a rule.

The last row is where most real business questions live. A query in Search Console tells you what someone wanted. A key event in GA4 tells you whether they got it. Neither half is useful without the other, and that is why we built ONE Dashboard to show both in plain English with read-only access. Take care with the word "conversions" in any report, though. Ours were wrong by a factor of nearly 50 until we looked closely.

Stop reconciling, start reading direction

You will never make these two numbers agree, and hours spent trying are hours taken from fixing things. Learn the size of your normal gap, watch for the moment it changes, and let each tool answer the question it was built for.

See your Search Console and GA4 numbers on one page and find your normal gap: connect your site to ONE Dashboard free.