The last touch didn’t close the deal. Stop crediting it.

Having worked across both B2C and B2B, the difference in how each world thinks about attribution is stark. In B2C, the infrastructure is mature. The feedback loop is tight, the buyer is one person, and the tools were built specifically for that journey. You could measure, test, and optimise with reasonable confidence that what you were measuring was real.

Moving into B2B, most marketers bring those same tools with them. The same models, the same reporting logic, the same instinct to find the touchpoint that closed it. And for a while, nothing seems wrong. The dashboards still populate. The attribution reports still get presented in quarterly reviews. The numbers still go somewhere.

The problem is what the numbers are actually measuring.

What B2C attribution taught B2B the wrong lessons

B2C attribution matured quickly because the feedback loop is tight. A customer converts within hours or days, the touchpoints are finite, and you can run controlled experiments at scale. The infrastructure built around that reality is genuinely sophisticated.

B2B borrowed it anyway. Not because it fit, but because it was there.

In B2B, a deal closes 3 to 18 months after first contact, and buying committees involve 3 to 12 people, each engaging with different content on different timelines (Improvado, B2B Marketing Attribution Models, 2026). Some of those people will never appear in your CRM at all. Peer referrals, conference conversations, a mention in a private Slack community, none of it shows up in a dashboard. The same research estimates that between 38 and 51 percent of B2B pipeline is driven by these untrackable interactions — what the industry now calls the dark funnel.

That is not a rounding error. That is most of what is actually closing your deals, sitting completely outside any model you have access to.

The last touch problem

Last-touch attribution gives 100 percent of the credit for a closed deal to the final recorded interaction before conversion. In B2C, where the journey is short and the buyer is one person, this is imprecise but defensible. In B2B, it is close to fiction.

What last-touch typically captures in a long sales cycle is the administrative moment, the demo request, the contract link click, the final check-in email. Not the blog post a technical evaluator read eight months ago. Not the LinkedIn content that made the CFO comfortable enough to agree to a first call. Not the case study your champion printed and brought into the committee room. All of that work registers as zero.

The result is that marketing budgets get allocated toward whatever activity sits closest to conversion, and whatever built the conditions for that conversion gets quietly defunded.

Why nearly 90 percent of B2B teams are still doing this

The industry is aware of the problem. RevSure’s State of B2B Marketing Attribution report (2025), based on interviews with over 60 senior B2B marketers, found that close to 90 percent of teams are still relying on single-touch or basic multi-touch attribution models, despite widespread acknowledgment that these models do not reflect how B2B deals actually close. SegmentStream’s 2026 B2B attribution guide corroborates this, noting that last-click attribution and platform-reported metrics remain “fundamentally unreliable for B2B decision-making” yet continue to drive budget allocation across most revenue teams.

The reason is not ignorance. It is that the alternative is harder to defend in a boardroom. A last-touch report is clean, causal, and easy to present. A more honest account of how a deal developed involves a chain of interactions that cannot all be tracked, a committee whose individual journeys are partially invisible, and a significant portion of influence that happened entirely offline. That story is true. It is also uncomfortable to put in a slide.

So the cleaner story wins, and budget decisions get made on top of it.

What resolution actually looks like

The companies reducing customer acquisition costs and improving ROI are not the ones who found a perfect attribution model. They are the ones who stopped looking for one. Improvado’s 2026 research found that teams achieving 15 to 30 percent CAC reduction were not using more sophisticated models — they were combining sufficient modelling with incrementality testing and qualitative data to fill the gaps their tools could not.

What that looks like in practice: asking closed-won customers directly how they found you, what influenced the decision, and who else was involved. Treating that data with the same seriousness as anything in the CRM. Accepting that a portion of pipeline will always come from interactions that leave no digital trace, and building that assumption into how performance gets evaluated rather than pretending it does not exist.

Attribution in B2B is not a measurement problem you can solve with a better tool. It is a maturity problem that requires the business to agree on what it is actually trying to understand, and then build reporting honest enough to say so.

The last touch closed nothing. It just happened to be standing there when the deal arrived.