The pipeline question you can't quite answer

Picture the question arriving the way it usually does: late in a quarterly review, almost as an aside. The CFO has been looking at the event line. "What did the dinner series actually produce?"

You have an answer, and you believe it. Say you watched a VP from a stalled account spend two hours talking with your CTO, and the deal moved the following month. What you do not have is a number that survives the next question.

This is the most familiar anxiety in field marketing, and the research explains why it refuses to go away.

Two numbers that refuse to meet

Start with what marketers believe. In Splash's 2025 survey of 1,058 US marketers, 52% said events drove at least half of their company's 2024 closed-won deals (Splash, 2025). Splash sells event software, so this is vendor research, but the sample is large and the belief is common.

Now look at what the CRM records. HockeyStack, an attribution vendor, analyzed 2.64 million deals across 198 B2B SaaS companies. Under linear attribution, events combined accounted for about 6.5% of closed deals (HockeyStack, 2025).

These are not the same measurement. The samples differ, the methods differ, and one number is self-reported while the other is derived from systems. Set side by side, they are a contrast, not a comparison, and anyone who divides one by the other to claim marketers overstate events by some multiple is doing bad math.

The distance between them is still the thing that keeps field marketers awake. One number is what the people closest to the work think happened. The other is what the system of record can see. Your CFO is going to believe the second one.

Why both numbers are probably wrong

The linear figure has a structural flaw. The average B2B buyer journey runs 211 days and involves 76 touches, 6.8 buyer stakeholders and 3.7 channels (Dreamdata, 2025). A linear model splits credit evenly across every recorded touch. A dinner that changed the direction of a deal gets the same weight as an email open, and a dinner that was never logged gets nothing.

Timing compounds it. Buyers make first contact with sellers about 61% of the way through the buyer journey on average, according to vendor research from 6sense (6sense, 2025). Much of what shapes a shortlist happens before any trackable touch exists. If your event sits in that early stretch, a click-based model will struggle to see it.

The self-reported figure has the opposite flaw. It comes from people whose budgets depend on events looking productive, answering a question about influence that has no agreed definition. "Attributed to events" can mean anything from "sourced by an event" to "someone from the account once stopped by the booth."

So the honest position is uncomfortable. The real number is very likely above what linear attribution shows and very likely below what marketers feel. Nobody can say where, because most teams are not collecting the data that would settle it.

The data most teams skip

Here the anxiety turns out to be partly self-inflicted. More than half of marketers do not track opportunities created as a measure of event effectiveness (Splash, 2024). In Forrester's 2026 survey, 44% of B2B organizations measure event impact, and 30% struggle to show impact even with metrics in hand (Forrester, 2026).

The pressure above the event team keeps building. Only 52% of senior marketing leaders say they can prove marketing's value and get credit for it (Gartner, 2024). 47% of CMOs say marketing is perceived as an expense rather than a strategic investment, and CFOs are the most skeptical executives in the building, at 40% (Gartner, 2024). When the CMO cannot win that argument, the event line is an easy place to find savings.

There is one modest piece of encouraging evidence in the HockeyStack dataset. Deals touched by live events converted from created to qualified at 5.50%, compared with 4.82% for other channels, and averaged a 12.1% monthly created-to-closed-won rate against 11.1% (HockeyStack, 2025). The gap is small and vendor-reported, and it does not prove cause. But it points toward a more defensible kind of claim than "events drove half our revenue."

What to do before the question comes back

Winning the attribution argument is the wrong goal. Replacing a belief with evidence you can show line by line is the right one.

Pick a claim you can defend. "Sourced" is a fight you will lose. "Influenced" without a definition is one you will also lose. A narrower claim holds: of the accounts with open opportunities, these ones had a contact attend, and here are their names. That is a list, not a model, and a CFO can check it.

Measure at the account level. A typical buying decision now involves 13 internal stakeholders and 9 external influencers (Forrester, 2026). A contact-level lead score cannot tell you whether the right people from an account were in the room. An account-level view can.

Compare your own cohorts, carefully. Pull opportunities where someone attended an event and opportunities where no one did, and compare how they progressed through stages. Say out loud that the groups are not random, since you invited the warmer accounts. The comparison is still more persuasive than an industry survey, because it is your data.

Get on the calendar with analytics. Gartner found that 62% of marketing leaders who meet regularly with analytics leaders prove value and get credit, versus 30% of those who meet infrequently (Gartner, 2024). For a field marketer, that usually means a standing slot with RevOps before the event, to agree on what will count, rather than after, when the numbers are already disputed.

Log the attended list the next morning. Every week it sits in a spreadsheet, it gets harder to match to the CRM and easier to dismiss.

None of this produces a single, satisfying ROI figure. It produces an answer to the CFO's question that is specific enough to check, which is usually what the question was asking for in the first place.