DZ_Blog_B2B_ElephantFiles_E2_web

The Elephant Files: B2B marketing attribution and the number nobody believes.

Published on 16 November, 2026 | Author: Agnes Marggs | 8 min read

Somewhere in your company there is a contract that no one ever signed. It has no letterhead, no legal review, and it never reaches the data room. Even so, in B2B companies everywhere, the same three parties renew it every quarter without saying a word of it aloud. Marketing agrees to bring a number to the review. Finance agrees not to ask too closely how that number was built. Sales agrees to hold on to the right to wave it away when a deal goes sideways. And all three agree to leave one question unasked: how much of the buying journey behind that number could anyone actually see?

That contract is Elephant 2, the second open secret in this series, and it lives inside the work we call B2B marketing attribution. The Elephant Files exists to bring these conversations into the open, the ones marketers trade over coffee and rarely repeat in the room where budgets are decided. The first file followed the money, the spend that kept climbing while AI promised savings. This one follows the credit, the proof of impact that marketing is so rarely able to show.

It would be easy, and wrong, to read this as marketers inventing numbers. They do not. The figure is produced by a system: an attribution tool runs its model, assigns credit to whatever it managed to track, and returns something like “marketing sourced $4.2 million.” That figure is real, in the narrow sense that the software genuinely generated it. What no one can vouch for is the certainty wrapped around it, because the model only ever sees a sliver of a long, mostly invisible journey. Everyone in the room knows this much. The quiet agreement is to present the figure as settled anyway.

Which raises the question the contract exists to bury: if a number is incomplete, and everyone present knows it, why does it survive every budget review intact? The answer sits in the terms, read clause by clause.

Clause one: a number is non-negotiable.

The pressure to bring one is real, and it keeps rising. The CMO Survey’s Spring 2025 edition found that demonstrating marketing’s impact on financial outcomes is now the single hardest challenge marketing leaders face. Across the same period, the share of marketers reporting heavier pressure from the CFO rose to 63%, and pressure from the board reached half of all respondents. The budget will not stretch to absorb that scrutiny. When leadership wants proof, a defensible number quietly outranks an honest “we can’t fully say.” That is the first term of the contract, and it is the one nobody resents.

Tabitha Naylor, a fractional marketing leader with more than 20 years building demand generation and lifecycle systems for B2B organizations, describes the number that actually reaches the boardroom. “I believe the direction. I don’t believe the decimal,” she said. “The number I present is usually the number that survives the meeting, not the whole truth.”

Clause two: the number must be defensible, not true.

Here is the uncomfortable part few will say at the table: most of the people presenting these numbers do not fully believe them. 6sense’s 2025 research on B2B marketing metrics found that while most teams now run an attribution model, only about one in five uses the kind of statistical model that could reflect a complete buying journey. The popular multi-touch models credit three or four visible touches, a rounding error against the dozens of interactions a buying group actually has. The doubt is not only external. A 2025 TransUnion study found that 60% of marketers say their own internal stakeholders question the validity of their metrics, and nearly three in ten say up to a fifth of their budget has been reallocated or put at risk over those doubts. If the people closest to the number hesitate over it, why does it still anchor the quarterly story? Because a defensible figure keeps the conversation moving, and an honest hedge stalls it.

Aimen Hallou, chief technology officer at Floxy, has seen the gap directly in his own funnel. “B2B multi-touch attribution dashboards distort paid digital channel contributions by 30-50%, at the expense of dark social and word of mouth,” he said. “When we evaluated our inbound enterprise funnel, multi-touch software credited 45% of closed-won revenue to paid search ads, while post-purchase surveys found 68% of clients actually came through word of mouth in private Slack channels. The software caught the last click and handed it credit for brand reputation it never built.”

Clause three: the real origin stays off the record.

The reason the number is incomplete is simpler than the dashboards suggest: most of the buying journey happens where no tool can watch. Gartner’s research shows B2B buyers spend only about 17% of the journey with sales reps, and closer to 5% with any single vendor’s rep. Its 2026 surveys put the preference for a rep-free experience at 67%, with 70% wanting to buy through fully self-serve channels. By the time a buyer makes contact, the decision is largely set. 6sense’s 2025 Buyer Experience Report finds the eventual winner sits on the buyer’s Day One shortlist 95% of the time, and the pre-contact favorite goes on to win roughly four of every five deals. Dreamdata’s 2026 benchmarks, drawn from millions of real journeys, trace an average of 88 touchpoints across 10 stakeholders, with 81% of the journey unfolding before the deal ever reaches the pipeline. Where, then, do the best deals actually come from? Mostly from conversations the dashboard never recorded, which is why it quietly hands the credit to the last click it happened to catch.

Kerry Cunningham, head of research and thought leadership at 6sense, has watched marketers misread his own data the same way. “Marketers assume their future customers need search engines or LLMs to discover them,” he said. “What the data actually shows is that about 75% of buyers, in most categories, have known the vendors for years. They’ve evaluated them and used them many times before. Their buying journey starts in memory, long before it starts in a search bar.”

Clause four: sales keeps the right to disown the lead.

The final party signs for the safest terms of all. The same 6sense metrics research notes that the MQL most teams claim to have retired still quietly governs their reporting and their compensation. The credit rule that follows is lopsided by design. A number that looks good becomes shared success across marketing and sales. A deal that dies becomes a bad lead. The arrangement shields all three parties at once, which is precisely why none of them reaches for the pen to cancel it.

Why a better metric won’t void the contract.

The fashionable remedy is to retire the MQL and measure pipeline influence instead, a direction Digitalzone itself has named the defining shift of 2026. The direction is right, and it is still not the finish line. Trading a weak metric for a stronger one leaves a number you cannot fully prove, because the journey beneath it remains largely unseen. A better metric makes the contract more sophisticated. It does not tear it up. The honesty problem was never about which metric you choose. It is settled only when someone says, in plain words, what the number can and cannot tell you.

Rewriting the contract.

What would an honest number look like? It would carry a range rather than a false decimal, and label plainly which part is measured and which is inferred. It would treat self-reported attribution as real evidence, the unglamorous “how did you hear about us” answer that surfaces the dark funnel no dashboard can see. It would fund the invisible middle on purpose, the brand and content work that shapes a shortlist months before a form is ever filled. The teams that win the budget argument in 2026 will not be the ones with the most confident dashboard. They will be the ones who can state, without flinching, exactly what their number proves and what it does not.

Chris Walker, founder of Refine Labs and one of the loudest B2B advocates for asking buyers directly, draws the line in a recent interview. Self-reported attribution, he said, typically tells you “what created the demand, not what captured the demand.” That is the half of the story the dashboard was never built to see.

The case file stays open.

The next time the quarterly number lands on the screen, watch the pause around it. That pause is the contract, renewed for another quarter. You are not obliged to be the one who keeps quiet. Elephant 2 is named, and the room is filling up. 

Recommended Post

Digitalzone_blog_2026_Anti-AI Aerie campaign_web

What Aerie’s anti-AI campaign reveals about where marketing is headed

In a world of increasing inability to discern what’s real and what’s not, sometimes the most powerful thing we...
Digitalzone_blog_june2026__AI-disclosure-labels-on-creative_web

New research confirms AI disclosure labels have no negative impact on creative performance

As AI-generated advertising becomes standard practice, a new regulatory reality is taking shape alongside it....
Digitalzone_blog_2026_5.-10 emerging B2B marketing trends we're predicting for 2026_web

10 emerging B2B marketing trends we’re predicting for 2026.

The ground beneath B2B marketing keeps shifting. New technologies are pushing us into uncharted territory,...