Ask most B2B marketing leaders whether their attribution is solid, and you’ll get a confident yes. Ask to see the setup, and the confidence tends to evaporate.
I’m not trying to call anyone out; I just want to shine a light on a problem that usually gets swept under the rug.
Because I’ve been there when I asked: “How did we give attribution to the deal that touched six marketing tactics over four months before a rep closed it in a call nobody logged?”
Why does attribution stay broken?
Attribution problems don’t usually announce themselves. Revenue comes in. The pipeline is working. The dashboard still has numbers on it, and the numbers look like they mean something. What’s actually happening in most companies I’ve audited is closer to this: first-touch and last-touch models don’t get measured properly. Multi-touch attribution doesn’t get set up properly at a point in time, and then it never gets revisited as the buyer journey and channel mix potentially change month to month.
Nobody revisits it because it’s not broken in an obviously visible way. It’s broken in a way that we’re doing a ton of different marketing tactics and we’re not sure if we should keep doing all of them.
The three places I always check first
Where do sales-influenced deals get credited? In most B2B businesses, especially longer-cycle or higher-ACV ones, a meaningful chunk of deals are genuinely influenced by outbound, events, or relationship-driven sales motion that never gets logged as a “touch” in the CRM the way an email or lead gen form fill does. If your model only counts digital touches, you’re not measuring attribution. You’re measuring which channels are easiest to track.
Does your model match your actual sales cycle length? A 30-day attribution window on a nine-month enterprise sales cycle will systematically undercount early-funnel channels like content and organic, because by the time a deal closes, the tracking window has long since expired. This is one of the most common and most invisible errors I see, and it quietly biases budget toward bottom-funnel, high-intent channels every single time, but it’s not covering the entire content marketing picture.
Are “marketing-sourced” and “marketing-influenced” leads being reported as the same thing? They’re not, and conflating them is one of the fastest ways to either overstate or understate marketing’s contribution to the board, depending on which one gets used when it’s convenient.
So what’s the fix?
You don’t need a perfect model. Nobody has one. Even the most sophisticated marketing mix model setups are approximations, not ground truth. What you need is a model whose limitations you can actually understand and can state out loud, and a willingness to revisit it as your GTM motion changes.
The businesses I’ve seen get the most value out of fixing this are the ones who are doing a million and one content marketing tactics but could improve their lead generation by doing less, not more.
Why this matters more than it seems
Bad attribution doesn’t just make reporting look sloppy. It actively steers budget toward the wrong channels, quarter after quarter, because you’re optimizing against a systematically biased signal. I’ve seen companies cut a channel that was actually their best top-of-funnel performer, purely because the attribution model couldn’t see its contribution. That’s not a measurement inconvenience. That’s a strategic decision made on bad information.
If you haven’t stress-tested your attribution model against your actual sales cycle and actual sales motion in the last twelve months, that’s worth doing before your next budget cycle, not after.
If you want a second set of eyes on whether your attribution is telling you the truth, that’s exactly the kind of thing a marketing audit is built to surface. Schedule a call.
