Website traffic sources look like the simplest report in analytics: a list of channels with sessions beside each. People make budget decisions from it, and they often make them on a misreading. The channels are labels applied by rules, the rules depend on information the visit happens to carry, and visits that carry nothing end up in a bucket that looks like loyalty.
Google says as much in its own definition. In the Analytics help page on the default channel group, it describes channels as “rule-based definitions of your website’s traffic sources that let you monitor the performance of all of the channels sending traffic to your website.” Rule-based is the phrase to hold on to. The report does not know where a visitor came from; it knows what the visit told it, and it applies the first rule that fits. Change what the visit tells it and the same person lands in a different channel.
The channels GA4 sorts traffic into
GA4’s default channel group is longer than most people expect. It includes Direct, Organic Search, Paid Search, Referral, Organic Social, Email, Affiliates, Display, Organic Video, Paid Social, Paid Shopping, Cross-network, AI Assistant, Audio, Mobile Push Notifications, Paid Other, Organic Shopping and SMS. Anything that matches none of them goes to Unassigned. Most sites see only a handful of these with any volume, which is fine; the long list exists so that tagged traffic has somewhere precise to go.
The AI Assistant channel is worth noticing. Visits from AI tools now have a place of their own in the default grouping rather than being lumped into Referral. If you are working on visibility in those tools, our guide on how to rank in ChatGPT covers the other side of that traffic.
How the main website traffic sources are defined
The rules are specific, and reading them changes how you read the report. These are the definitions for the channels that cause the most confusion, quoted or summarised from Google’s default channel group documentation.
| Channel | Rule (per Google) | What it means in practice |
|---|---|---|
| Direct | “Source exactly matches ‘(direct)’ AND Medium is one of (‘(not set)’, ‘(none)’)” | The visit carried no referrer and no campaign tags |
| Organic Search | Source is on Google’s list of search sites, or medium exactly matches “organic” | Unpaid visits from a recognised search engine |
| Referral | “Medium is one of (‘referral’, ‘app’, or ‘link’)” | A link on another site or app that passed a referrer |
| Unassigned | Applied “when there are no other channel rules that match the event data” | The data exists but fits no rule, often from inconsistent tagging |
Notice what Direct actually requires. It is not “someone typed your URL”. It is the absence of a source and a medium. Typed visits and bookmarks qualify, but so does any visit where the referrer was lost on the way.
Direct is partly a measure of what you failed to track
The received wisdom is that a large Direct share signals a strong brand. Sometimes it does. But because Direct is defined by missing information, it also collects everything that arrived without a label: links clicked in some email clients and messaging apps, links in documents and PDFs, and visits from some mobile apps that do not pass a referrer.
Email is the classic case. A newsletter can send a burst of visits that, untagged, show up as a spike in Direct. Someone then credits brand awareness for traffic that the email team earned. The same happens with social posts shared through apps.
A quick way to test this on your own data is to look at what Direct visitors land on. Typed and bookmarked visits tend to arrive on the home page or a handful of well-known pages. If Direct sessions are landing on a deep article published the same morning you sent a newsletter about it, those visitors did not type that URL. Break Direct down by landing page and by date, and line the spikes up against your send schedule and social calendar. The overlap is usually obvious.
How to check website traffic source accuracy with UTM tags
The fix is boring and effective: tag every link you control. UTM parameters tell Analytics the source and medium directly, so the visit no longer depends on a referrer surviving. A newsletter link might look like this:
The medium matters most, because the channel rules key off it. Use the values the rules expect, such as “email” for email, and use them consistently. A team that writes “Email”, “e-mail” and “newsletter” in the medium field will scatter one channel across Email and Unassigned and wonder why the numbers never add up.
Keep a short shared list of approved source and medium values and check new campaigns against it. Do not tag internal links on your own site: a UTM on an internal link starts a new attribution and overwrites where the visitor really came from.
Reading a competitor’s traffic mix
Your own GA4 data is measured, with the caveats above. A competitor’s channel mix in a third-party tool is not measured at all; it is estimated from panels, clickstream and modelled search traffic. Those estimates can be useful for direction, and they can be badly wrong for any single site. Our analysis of how wrong traffic estimates are sets out where they break.
For your own search channel, Analytics and Search Console measure different things, and they will not match. Our comparison of Search Console vs Google Analytics explains which to trust for which question. And remember that many searches end without a click at all, which no channel report can see: our piece on zero-click searches covers that gap.
The limitation here is structural. GA4 can only classify information that reaches it. If a referrer is stripped, a consent banner blocks the tag, or a link was never tagged, no amount of report configuration recovers the true source. You can shrink the unknown share with disciplined tagging, but you cannot reduce it to zero, and any analysis that treats Direct as a clean brand metric is building on that unknown.
The practical routine is short. Tag everything you send, standardise your medium values, watch Unassigned as a tagging error report, and treat sudden swings in Direct as a question to investigate rather than good news to report. When you compare periods, check whether your tagging changed between them; a new tagging convention can move thousands of sessions between channels without a single extra visitor.
Frequently asked questions
What are the main website traffic sources in GA4?
The default channel group includes Direct, Organic Search, Paid Search, Referral, Organic Social, Email, Affiliates, Display, AI Assistant and several more, with Unassigned for anything that matches no rule.
Why is my Direct traffic so high?
Direct requires the source to be “(direct)” with no medium, so it collects any visit that lost its referrer, including untagged email and app links, as well as typed visits.
What does Unassigned mean?
Google applies it “when there are no other channel rules that match the event data.” It usually points to inconsistent or non-standard UTM values.
Can I see a competitor’s traffic sources?
Only as an estimate from a third-party tool. Those figures are modelled, not measured, and can be far off for an individual site.
The takeaway Channels are labels applied by rules, and Direct is what the rules return when a visit carries no information. Tag your links, standardise your values, and read Direct and Unassigned as gaps in tracking before you read them as results.
