Why last-click attribution misleads B2B teams
A B2B purchase is not a click. It is a sequence of visits spread over weeks or months, made by someone who is not ready to talk to you for most of that time. The visit that finally produces a demo request is the least interesting one in the sequence — by then, the decision to reach out has already been made.
Last-click attribution ignores everything before that final visit. And because of how buyers actually behave, the final visit has a predictable shape: it is Direct. The buyer read your comparison piece from a LinkedIn post in March, came back through a Google search in April, and in May typed your URL into the address bar because they already knew where they were going. A session-scoped report looks at that demo request and says: Direct converted.
The mechanism, precisely
Session-scoped reporting assigns each session its own traffic source, taken from the first hit of that session. The demo request lands in whatever session it happened in. If that session started without a referrer or a tagged URL, the key event is credited to Direct — regardless of the three channels that did the actual work. Multiply this across a quarter and Direct becomes your "best performing channel," which is another way of saying your reporting has stopped containing information.
The distortion is worst exactly where your spend decisions matter most. Paid search captures people who are already searching for a solution, so it converts in one session and looks efficient. Social and outbound create demand weeks before the conversion, so their credit leaks to Direct and organic search. The result: reporting that systematically tells you to fund the bottom of the funnel and starve the top.
What to measure instead
Three changes fix most of the picture. First, report key events against user-scoped, first-touch dimensions — the channel of the user's first-ever session — alongside the session-scoped view. The gap between the two columns is the size of your last-click distortion.
Second, look at paths, not points. Attribution path reports show the actual sequences: how often social or email appears early in journeys that end in a conversion, even when it never gets last-click credit.
Third — and this is the step most teams skip — close the loop. A demo request is not revenue. Until the qualified, worked, and closed stages from your CRM are written back against the original visitor, every attribution report ranks channels by their ability to generate form fills, not contracts. In our data, the channels that win on form fills and the channels that win on closed revenue are rarely the same list, and the ordering routinely flips once deal outcomes are attached.
Last-click is not wrong because the math is broken. It is wrong because it answers the question "what happened right before the form?" when the question that allocates budget is "what started the relationships that became revenue?" Those are different questions with different answers, and only one of them is worth a chart in a board deck.