How first user source works
When someone arrives on your site and GA4 has no existing record of them, it writes down where they came from and stores that value against the person, not the visit. Come back a week later through a Google search, a month later through a Facebook link, and the first user source stays what it was on the opening day. It is a permanent label attached to the visitor record.
The value itself is read from the referring site, the campaign parameters on the link, or the click identifier a paid platform appends to the URL. If none of those are present — someone types the address, taps a link inside a messaging app, or opens it from a PDF — GA4 has nothing to read and files the user under Direct.
Why first user source matters
It answers the question every business owner actually asks: where do new customers come from in the first place. Session-level reporting cannot answer that, because by the time a person enquires they have often returned through a branded search or a bookmark, and the channel that introduced them has disappeared from the report.
For a business with a long decision — an education consultancy, a clinic, a remittance service, a builder quoting on a project — the gap between the first visit and the enquiry can run into weeks. Without a user-scoped source, everything that started on social or display quietly gets credited to organic search or direct, and budgets follow the wrong channel.
Common mistakes with first user source
The most common one is putting it next to session source in the same table and treating the difference as a tracking fault. They are different scopes measuring different things, and a mismatch is expected rather than broken.
The second is trusting it after a tracking change. First user source is written once, so if campaign links went out without UTM parameters, or the analytics tag was missing from the landing pages for a month, those users carry the wrong label for as long as their record survives. Cookie expiry, browsers clearing storage and people switching between phone and laptop all manufacture new “first” visits too, so treat this as a strong indicator rather than a census.
How to act on it
Tag every link you control — ads, email, social bios, QR codes on print — so there is something for GA4 to read. Then judge acquisition channels on user-scoped reports and judge on-site behaviour on session-scoped ones, rather than mixing the two in a single view and arguing about which is right.
If the direct bucket is large, treat it as a tagging job before treating it as a real channel: check that WhatsApp and Viber links, app links and offline campaigns all carry parameters. In Nepal, where a great deal of traffic arrives through a message rather than a browser search, that one fix usually changes the picture more than any amount of report configuration. A structured GA4 setup is what makes the dimension trustworthy in the first place.