I agree that's the right metric, but I doubt that the mechanism is actually application volume/spam/low-quality applications.
For instance, the number of days a role is open is *also* flat on top of the flat time from application-to-hire. It just doesn't appear to be true that matching is the crux
