You would at least be able to act on the true cause rather than chase short term changes that may not even work or won't scale. If it's indicative of a distribution problem, then we should be investigating distribution solutions. If you can't see this connection, then I posit that you might have the shallow perspective.