I think you're confusing "virality" with "popularity". Or maybe it's a God of the Gaps argument -- if we looked at sufficiently popular stuff we'd see viral pro

I think you're confusing "virality" with "popularity". Or maybe it's a God of the Gaps argument -- if we looked at sufficiently popular stuff we'd see viral propagation! The study considered a pretty diverse set of online stuff, with, I presume, many millions of total adoptions. Essentially none of those adoptions were due to viral propagation. That seems compelling to me. (PS, I'm Sharad Goel's coblogger at Messy Matters though I wasn't involved in this research.)