In reality, of course, there probably just isn’t that much money at play so the high-budget approach is unlikely to break even, and more critically it’s unlikely that people are prepared to change how the organization makes decisions. I’ve known some people who worked for places where they had data showing solid improvements but it either hit politics or was simply small enough that their executives didn’t want to go with it because it wasn’t the magnitude they’d promised to justify the data science program.
