Your pipeline meeting is a translation problem
Friday, ten a.m., pipeline review. Marketing presents 340 MQLs for the quarter, 112 percent of target, slide with a green checkmark. Sales looks at its own list and finds roughly 60 that deserve the name. The rest: whitepaper downloads from students, competitors with throwaway addresses, an intern who filled out the same form three times. And then comes the sentence that gets said at every other midsize company. Your leads are garbage. Followed by the answer that’s just as old: you don’t even touch them, four days on average before first contact.
Both sides are right. That’s the problem.
Marketing counts anything with a score above 50 as an MQL. Sales means a contact with budget and a concrete project. And if you have a CS department, they probably use the word for something else entirely. Three departments, three funnel definitions, one term. As long as that’s the case, every meeting is a discussion of numbers that don’t measure the same thing, so naturally you argue about the numbers instead of about actions. The meeting feels like work. It’s simultaneous interpreting without an interpreter.
The reflex then is often to buy a tool. Better scoring, a new attribution model, a dashboard that finally shows the truth. Doesn’t work, because the problem isn’t a data problem at all. Unified funnel definitions are a decision, not software. Someone with authority over both departments has to rule: an MQL is a contact from the ICP with behavior X, an SQL exists once sales has confirmed criterion Y, and the handoff only counts after a documented first touch. That’s an afternoon of work plus one uncomfortable conversation. No license agreement in the world does that for you.
After that, and only after that, the dashboard is worth it. Because a dashboard on inconsistent definitions isn’t reporting, it’s an argument accelerator with charts. The same conversion rate, calculated three different ways, interpreted three different ways. We’ve seen companies pour 40,000 euros into BI tooling before anyone had written down what a lead actually is.
The test is simple. Ask three people from three departments what an MQL is at your company, in writing, without letting them coordinate. If three different answers come back, you know why your forecast wobbles. And you also know the fix costs nothing except a decision nobody wanted to make until now.
