Nobody invented the marketing qualified lead because it predicted revenue. We invented it because we could not see revenue, and a function that cannot see its own outcome will always build a proxy it can count. The MQL was a reasonable answer to a real problem in 2009. It is now the primary reason marketing is the last department in the building trusted with a number.
Here is the mechanism. A proxy starts as an approximation of the thing you want. Then it becomes the thing you are measured on. Then it becomes the thing you optimize. By the third step it has quietly detached from the outcome and the optimization is working against you. You can watch this happen in any funnel review. Lead volume goes up and to the right, sales says the leads are garbage, marketing produces a slide proving the leads are not garbage, and the meeting ends with a request for more leads.
A metric both teams distrust is not a bad metric. It is a symptom that the two teams are being paid to want different things.
The standard fix is to raise the bar. Add fit scoring, add intent data, rename it an SQL, attach a service level agreement about response time. This is tightening the strap on a broken instrument. It does not matter how carefully you define a lead while the definition is owned by the team rewarded for producing more of them.
The actual fix is uncomfortable, which is why it is rare. Marketing has to be measured on marketing sourced revenue and nothing else. Not influenced revenue, which is a number designed to be unfalsifiable. Sourced. The pipeline that would not exist without us, tracked to close, reported at the same cadence as the sales number and in the same meeting. Everything else, cost per acquisition, lead volume, campaign efficiency, channel mix, becomes what it should always have been, which is diagnostic. You look at it when the revenue number moves and you want to know why.
This is a worse metric in exactly one way. It is slow. In a business with a six month cycle you will spend two quarters unable to prove you are working. Teams cannot tolerate that ambiguity, so they reach for something faster, and the something faster is always a proxy, and we are back at the beginning. The discipline required is not analytical. It is political. Somebody senior has to accept being judged on a lagging number and then hold the line while it lags.
I have run this both ways. The teams where I insisted on sourced revenue were harder to manage in month two and unrecognizable by month nine, because every argument about tactics resolved against a single number nobody could game. The teams where I let the MQL survive spent their energy defending lead quality instead. Same people. Same budget. Different instrument.
If you are hiring a marketing leader right now, ask what number they want to be measured on. Then ask what happens in month three when it has not moved yet. Anyone can name the right metric, it is in every article about marketing leadership. The second question cannot be answered from reading, because it requires having sat in a room where a lagging number had not moved, the board was uneasy, and somebody had to either hold the line or start manufacturing something faster to show.