Last Non-Direct Click

Last non-direct click is an attribution rule that skips direct visits when assigning credit. A direct visit is one where the person typed your address, used a bookmark, or arrived without a referrer your analytics tool could read. Under this rule, if a customer clicks an email on Monday, comes back by typing your URL on Wednesday, and buys, the email gets the credit, not the direct visit. In Google Analytics 4, the help page says that 'paid and organic last click' and 'last non-direct click' are two names for the same attribution model, and that it gives full credit to the last channel the customer clicked before converting. The idea is easiest to see on a returning customer: someone who clicked an ad last week and types your address today did not discover you by typing it.

The logic rests on a reasonable guess: a direct visit is usually someone who already knows you, so the earlier source is the more useful thing to credit. The rule walks back through the path, discards direct sessions, and credits the most recent session with an identifiable source such as paid search, organic search, email, or a referral. If the whole path is direct, there is nothing to walk back to, so direct keeps the credit. This is why direct traffic often shows up large in reports: it holds every path with no recorded source. Direct is a catch-all bucket. Dark social, private messages, some apps, and links stripped of their referrer all land there, so it is a mix of loyal customers and lost information.

The rule sits inside a bigger set of GA4 choices. GA4 now offers three reporting attribution models: data-driven, paid and organic last click, and Google paid channels last click. It stopped offering first click, linear, time decay, and position-based in November 2023. Platform reports differ from GA4 reports, because Google Ads uses its own conversion credit and Meta uses its own windows. So last non-direct click in GA4 will not match what Google Ads or Meta report, and you should expect the three totals to disagree. GA4 also has a Google paid channels last click model, which credits only Google paid clicks, so the same path can show three different winners depending on the model you select.

The common mistake is trusting the rule to rescue untagged traffic. If your email links or social posts lack tracking parameters, they arrive as direct, and the rule treats them as direct, so they either vanish or get absorbed. Another is assuming it tells you who is responsible for a sale. It only tells you who was last among the sources your tools could identify. A third is using it for a channel mix with long gaps between visits, where earlier sessions may fall outside the lookback window. Consistent UTM tagging is the unglamorous fix for most of these problems, and it costs nothing.

Use it as a hygiene check on your tracking. Open the GA4 traffic acquisition report and note how much conversion credit sits in direct. A big share usually means missing campaign tags, links shared in private messages, or app traffic with no referrer. This week, audit your newsletter, social bio, and QR code links for tracking parameters, then re-check direct's share after a few weeks. If direct drops and email or social grow, your tagging was hiding real sources, and the model was working as designed. Compare your direct share month over month. A sudden jump often means a tracking change or a broken redirect, not a surge in loyal customers. Run it as a monthly habit rather than a one-time fix. New landing pages, new email tools, and new apps tend to introduce fresh untagged links, so direct creeps back up whenever tagging slips.