TL;DR: It really depends. Also: wrong question.
Conversion rates are interesting, people look at a percentage and take it as an absolute truth, it’s a number after all right?
What most operators miss tho is that every number has a story behind it, a series of assumptions and ideas that were baked into the calculation to get there. Not knowing what that story is more often than not sets you up for failure.
When it comes to MQLs, PQLs, and SQLs the funny thing is that definitions do matter: defining them in a different way will instantly increase or decrease the number of them that does qualify at any given point in time, with exactly the same number of Daily Active Users in the product and Monthly Website visitors on the website.
The other thing that most sophisticated operators always do is: beware of averages. Never ever trust them. In fact, I’d go as far as saying that averages are a lazy and misleading way to measure something, anything.
Given a representative enough base, you’ll most likely have a power-law distribution where you can identify at least 4 cohorts: slightly better performing, slightly worse performing, dramatically better performing, dramatically worse performing.
Breaking those cohorts down and thinking about them separately is already a way to inform the conversation in a significantly better way.
Would you want more “dramatically better” leads or should you optimize for more “dramatically worse” ones?