A good North Star Metric should reflect the value users get from a product, rather than just how often they use it. Airbnb, Facebook, and Slack are three interesting examples of how companies use metrics to guide growth, but each also shows the risks described by Goodhart’s Law.
Airbnb focused on nights booked because this number represents actual transactions between travelers and hosts. Unlike website visits or account registrations, bookings suggest that users find value in the platform. However, if Airbnb only tried to maximize bookings, it might encourage excessive discounts or misleading listings. The number could increase while customer satisfaction and trust decline.
Facebook’s early growth team identified seven friends in ten days as an important signal of user retention. The idea was that people who quickly built a social network were more likely to continue using Facebook. However, if Facebook pushed users to add as many friends as possible, people might connect with strangers or people they barely know. Friend counts would increase without necessarily creating meaningful relationships.
Slack used 2,000 messages sent by a team as an indicator of successful adoption. Reaching this number suggested that Slack had become part of a team’s daily communication workflow. However, optimizing only for message volume could encourage unnecessary conversations, constant notifications, and distractions. More messages would not necessarily mean better collaboration or higher productivity.
What I find interesting is that all three companies used observable behaviors as proxies for value. However, correlation does not always mean causation. Just because successful users reach a certain number does not mean forcing everyone to reach it will make them successful. I think product teams should combine North Star Metrics with measures like user satisfaction, retention, and productivity. Ultimately, the goal should be creating real value for users rather than simply making numbers look better.
