Free tool by Clarvivo · no login
Upload a GA4 traffic acquisition export. We compare how Direct behaves against every other channel and tell you whether it is real visitors, campaign traffic that lost its attribution, or automated traffic.
Export Reports → Acquisition → Traffic acquisition as CSV. We read channel, session and engagement columns only.
This tool answers: How much of my Direct traffic is real, and how much is lost attribution?
Enable JavaScript to run the check — the diagnosis itself is free and needs no account.
Every GA4 property has Direct traffic, and almost nobody knows what is in it. It is the channel of last resort: when GA4 cannot work out where a session came from, it goes to Direct.
Knowing the percentage does not help, because every site's baseline is different. A strong brand genuinely has a lot of Direct. A site with broken campaign tracking has a lot of Direct. The number alone cannot tell them apart.
What separates them is behaviour. Real direct visitors — bookmarks, type-ins, returning customers — are high intent and engage more than average. Lost attribution behaves like ordinary marketing traffic, because that is what it is. Bots barely engage at all. That is the comparison this tool makes.
These answer the questions this one tends to raise.
So you can trust the result — and your developer can verify it.
GA4 exports start with a comment preamble, name the same column differently across versions and locales, and format numbers with separators, currency symbols and durations like 1m 45s. We handle all of it, so you can upload the file exactly as GA4 gave it to you.
Direct is compared against a session-weighted engagement rate across every channel except Direct. Excluding Direct matters: including it would drag the baseline toward the very anomaly we are testing for, and a large enough problem would hide itself.
Direct engaging far below your baseline points at automated traffic — real people who typed your URL are more engaged than average, not less. Direct engaging at roughly your baseline, in large volume, points at lost attribution: ordinary marketing traffic wearing the wrong label.
If Direct earns a much larger share of revenue than of sessions, the channel that actually produced those sales is going uncredited — which is the version of this problem that changes budget decisions.
This export is one month, one property, one moment. Clarvivo joins acquisition data to payment data as it happens, so 'is this channel real' stops being a question you answer with a spreadsheet.
Replace attribution guesswork with Clarvivo