How does this site decide what to check for the use of a sample? We're at the point where sampling is so easy, so ubiquitous, that counting usage is impossible

How does this site decide what to check for the use of a sample? We're at the point where sampling is so easy, so ubiquitous, that counting usage is impossible without some arbitrary selection criteria (e.g. "I've heard of it", or, "The people who compile <some chart> have heard of it").