The lure of "the one metric that matters"

Description

Many products, whether data driven or not, chase “the one metric that matters.” It may be engagement, revenue, or some form of conversion, but the common theme is a single-minded pursuit of improvement in one metric. This approach to data science and product development leads to perverse incentives. Consider Facebook’s thirst for engagement being used against our democracy or Yahoo’s pursuit of revenue resulting in Yahoo! Mail turning into a digital Times Square. We data scientists, product managers, and engineers should spend time thinking about the design of metrics that measure our goals. This will make our products better and our customers happier.

Adil Aijaz shares an approach to designing metrics and discusses best practices and common pitfalls.