Data Science — A PM’s secret weapon

Distinction between a good PM and a great PM — the ability to use insights to influence decisions and execute.

I spent a few of my former years as a data scientist before transitioning into product management. My conclusion is that data science and product management are extremely similar in theory.

Both roles are in charge of solving a problem, whether it is a broader market problem or as small as optimizing a load balancer to increase capacity of your application by 10%.

The goal is keep a perspective but sweat the small stuff. For product managers, the small stuff is understanding constraints and benefits of releasing a metric driven feature. For data scientists, the small stuff might be to understand which activation function is best for your neural net.

Both roles require you to be analytical and experimental. As a product manager, having an understanding of experimentation in a broader business sense is just as complicated as running a structural time series bayesian model for a retrospective causal study on a single dataset.

In both roles, there is the setback of being too data driven. Having an intuition for getting a false positive or reaching a local max is just as important as releasing a product that’s good for short term revenue but bad for the brand.