There is no right vs wrong way of doing things, only how they are being done today, how we are learning from that, and how we decide to do them tomorrow.
Productive iteration is a key component of data maturity, and here’s how and why you should do it.
Bryan McCann shows how we teach machines to learn quickly from “mistakes” in AI; taking a page out of a machine’s playbook facilitates constant improvement and ultimately winning.
So-called “failures” are opportunities to improve our models with more data+outcomes to move forward with; remember what Giannis Antentekumopo said about ego vs pride? Those who leverage these productively will prevail. Calculated risk-taking should be encouraged, given that innovation is everyone’s job.
Charlie Munger talks about holding two conflicting ideas in your head at once and challenging yourself to take the other side of an argument to be sure you’ve really thought things through.
Data-driven insights can and should go hand-in-hand with human intuition and experiential, subjective inputs.
The traditional line that what can be measured, can be managed has been challenged by data points per athlete per year increasing 40x since 2000.
Just because you CAN measure it, doesn’t mean it matters; nobody who’s been inside a locker room would say that football teams should go for it on fourth down every time! This dynamic applies across analytics.
Productive iteration can be initiated by focusing on a discrete data set that you haven’t yet actioned but have a high degree of confidence in projecting the outcome generated by the traditional approach. This time, run an innovation sprint or two and develop 2+ prototypes. See what happens. Iterate. Do it again.