You are right on two counts. First that i should have said correlated rather than proportional. Second that even if there's an inverse influence, there's also

You are right on two counts. First that i should have said correlated rather than proportional. Second that even if there's an inverse influence, there's also a positive influence between both forms of advertisement. But in terms of proportion I still maintain that if you eliminate one type of advertisement the ratio will become 100% of the other, which is as undeniable as it is tautological.