We just wrapped a two-and-a-half-day event with a group of senior B2B marketers from companies you’d recognize. The format was discussion-based by design. No keynotes. No polished presentations. Just real conversations about what’s actually happening in marketing right now.
I walked in with data. I walked out with something harder to quantify: confirmation that the thing we’ve been calling the “confidence gap” is more pervasive, more personal, and more urgent than even our own research suggested.
Here’s the finding that keeps sticking with me: more than half of the marketing teams we surveyed described their content strategy as “advanced.” I want to be fair to that number. Advanced is a loose word. But when you hold that against what the same teams report about measurement confidence, pipeline contribution, and data quality, something breaks.
Nearly two-thirds say they don’t have meaningful confidence in how they attribute results. Fewer than one in five list pipeline as a top KPI. Most teams are measuring paid viewers, followers, click-through, and organic traffic. None of those are wrong, exactly. But if you had to defend your budget to a CFO with only those numbers, you’d be in trouble.
What I keep coming back to is that this isn’t a capability problem. It’s a systems and incentives problem.
One of the conversations that landed hardest was about Goodhart’s Law. If you’re not familiar: once a measure becomes a target, it stops being a good measure. In marketing, this plays out constantly.
Someone in the room told a story about a company where the CMO, in an effort to create clarity, set a single KPI: scheduled meetings. Clean, simple, trackable. So what happened? The team started paying customers fifty dollars to take meetings. They hit their number. They did not grow their business.
Nobody in that room was surprised. Because most of us have seen versions of this. The attribution model that makes one channel look good, so you over-index on it. The MQL target that gets gamed the moment reps figure out the scoring rules. The vanity metrics that look great in a board slide and mean almost nothing.
The problem isn’t that these things are malicious. It’s that they’re rational responses to the wrong incentive structure. You measure what you’re told to measure, you optimize for it, and somewhere in that process the actual goal gets lost.
The harder conversation, the one I don’t think gets said out loud enough, is this: a lot of marketing teams know their data is bad. They know their attribution is imperfect. They know their dashboards show activity, not momentum. And they’ve made a quiet, understandable decision to live with it because fixing it would require political capital, budget, and time they don’t have.
I’ve been in that spot. I’ve had the CRM data that I knew was garbage and no one above me seemed to care enough to fund the cleanup. I’ve run campaigns and then argued in retrospect about what they influenced because the systems couldn’t actually tell me.
What I saw in these conversations was a lot of people who are smart, experienced, and genuinely trying to do the right thing, held back by foundational issues they can’t solve alone. That’s not a comforting conclusion. But I think pretending otherwise is part of the problem.
There’s something genuinely difficult about the current moment for B2B marketers. AI has made it easier to produce more content, run more campaigns, and fill more dashboards with data. But producing more is not the same as improving signal. If anything, the volume increase makes the measurement problem worse, not better.
The teams that came out of this event with something useful weren’t the ones who had the most tools or the biggest budgets. They were the ones who had decided, at some level, to slow down and get deliberate: about what they were measuring, about what they trusted, about where they were willing to say “we don’t actually know.”
That last part is harder than it sounds. But I think it’s where the real work starts.
We recently released the full Confidence Gap report. If you want to benchmark your team’s maturity across measurement, data quality, and AI adoption strategy, it’s worth a read. Not because the data is going to tell you something you don’t already suspect, but because sometimes having the number makes the conversation easier to have.
More to come.