When Desai's team first reached out to business units in search of questions where data science might be useful, he did not require them to utilize "big data" (i.e. look for patterns to explore in the data). Why not? Do you agree?