The MQL was invented so marketing had something to commit to in a board meeting. It has outlived its usefulness by about 15 years. At this point, "we hit our MQL target" is the B2B equivalent of "we had a strong quarter of strategic alignment."
I do not measure MQLs on client engagements. When I walk in and the CRO is forecasting off MQL-to-SQL conversion rates, the first thing I change is the metric. Not because MQLs are dishonest. Because they are uselessly far from revenue.
Here is what I replace them with.
An MQL is a definition. That is its whole nature. "A lead that hit a behavioral or firmographic threshold defined by marketing, handed to sales." The definition itself is the product.
Which means: any time the number is bad, someone can change the definition. Raise the lead score threshold. Add an industry filter. Remove the ones from countries you do not sell into. The number goes up. Nobody sold anything new.
The reverse is also true. If marketing needs a big number for the board, lower the threshold. The number goes up. Still nobody sold anything new.
An MQL target is a target you can hit without affecting revenue at all. That is not a leading indicator. That is a vanity number with a SLA attached.
Every engagement gets the same replacement metric: sales-accepted opportunities created per week, by ICP segment.
Three things matter about that phrase. I will take them in order.
Sales-accepted means the rep actually picked up the opportunity and started working it. Not "received an email." Not "scored above a threshold." Picked up. Worked. This is verified against the CRM, not a marketing automation platform.
Opportunities created per week means weekly, not monthly. Weekly numbers move. Monthly numbers are postmortems. If the team cannot see the number change this week, the team cannot change its behavior this week.
By ICP segment means the number is decomposed. Not "we created 42 opportunities." But "we created 18 in ICP-1, 14 in ICP-2, 6 in ICP-3, 4 in unsegmented." The moment you decompose, it becomes obvious where the pipeline is actually coming from, and almost always, one of the ICPs is carrying the other two.
This metric is hard to game. You cannot change the definition of "rep picked it up." Either the CRM says they did or it did not. You cannot change "opportunity created" without the rep objecting — the activity data is tracked whether you like it or not. Decomposing by ICP means even if the aggregate number looks fine, a segment collapse becomes visible immediately.
It is also closer to revenue. An opportunity has a stage. A stage has a progression model. A progression model has a conversion rate. You can see the forecast implied by this week's number in about 10 seconds. You cannot do that with an MQL.
And it gives marketing something real to commit to. "We will create 20 ICP-1 sales-accepted opportunities per week for the next 10 weeks" is a commitment. "We will hit 800 MQLs this quarter" is a number.
Every time I replace MQLs, I get the same three objections.
"But we have attribution data on MQLs." You have attribution data on every touchpoint already. The MQL definition is not what enables attribution. Your data model is.
"But the marketing team will look bad." The marketing team is not looking good right now, they are looking safe. Different thing. When marketing commits to a metric the CRO actually forecasts against, marketing's status inside the company goes up. Marketing becomes a revenue driver, not a cost center running vanity pipelines.
"But that is just SQLs." Close. SQL conventionally means "qualified by sales after discovery call." I mean the step before that — rep picks up the inbound or outbound, starts working it, the opportunity is in the funnel. Earlier than SQL, harder than MQL.
One chart. Weekly. Stacked bars by ICP segment. Horizontal line at the commit number. Red when the line is not hit two weeks in a row. Black when it is.
That chart is the weekly review. That chart is the leading indicator. That chart is the conversation marketing and sales should be having.
If the chart is fine, we talk about what to scale. If the chart is broken, we talk about which segment is leaking, which source is drying up, and what is changing by Monday.
That is the whole stack. No funnel deck. No MQL-to-SQL handoff workshop. No "alignment" anything.
It is less fun for consultants. It is better for the business.