Finding talent is my largest challenge. Someone who understands our business, who has quantitative skills, who has the technical skills to create the models, and who is able to persuade others that the insights they’ve come up with are ones you can trust and take action on. The hardest part is persuasion. You get the quantitative skills, but there’s a struggle in that ability to communicate effectively. We’ll often pair people together, but we’d really like to grow the talent.
When I was in marketing, we put a focus on liberal-arts-educated individuals, because abstract thinking where there are ambiguous data sets is an area where they are comfortable. Ph.D.s in psychology were a great recruiting pool. A psych Ph.D. has a fair amount of statistical training. We created a program to recruit Ph.D.s.
There’s not yet an educational discipline and curriculum that produces data scientists at the scale that would clear the market. So the way we’ve focused on it is to find people with innate curiosity and critical thinking. You can teach the other skills. On my team, I have a pathologist, a bioengineering student who trained in doing heart research, an M.B.A., and someone who is trained in traditional data architecture. I also have a landscape construction engineer and a psychology Ph.D.