"Video views" are not a good measure of church attendance. "People reached" is even less accurate than video views when it comes to church attendance. So what do we do?