I think my only issue with your take is that you are treating a decline in what you can see as a decline for each person. I think you're probably seeing a result of lower filtering / more access rather than a decline within a consistent group