5% of encounters != 5% of people. In fact, even if the poster tried marketing to the same people at different times, he might well have gotten a different 5%. S

5% of encounters != 5% of people. In fact, even if the poster tried marketing to the same people at different times, he might well have gotten a different 5%. Some correlation, for sure, but still. Not to mention how encounters are a binary thing - another person might get a different 5%. Also, the samples he's talking about are not really uniformly distributed over the world's population (i.e. "people"), nor even people in the USA.