Could you share what kind of data points you use to describe "broken and convoluted processes"?