Teams with fifteen metrics on their dashboard have no shared understanding of what success means. When everything is measured, nothing is prioritized. The north star metric solves this by creating one number that the whole team — engineering, design, marketing, sales — can orient around. The debates shift from "which metric should we optimize?" to "how do we move this specific number?", which is a much more productive kind of disagreement. Having too many metrics is functionally the same problem as having none: both leave the team without a clear answer to "are we winning?"
Why north star metrics should measure value, not revenue
Revenue is a lagging indicator of value delivered. If you optimize directly for revenue, you create incentives for dark patterns, upselling, and retention at the expense of user satisfaction. If you optimize for value delivered — nights booked, listening hours, messages sent — revenue follows. The north star metric should be the number that, if it goes up, you are confident the business is healthier, even before you see the revenue impact. This distinction is not just philosophical. Teams that optimize for revenue directly tend to ship features that extract value from users. Teams that optimize for value delivered tend to ship features that create it.
What the metric tree reveals that the north star alone hides
The north star tells you whether things are going well or poorly. The metric tree tells you why. A metric tree that breaks DAU into new users, retained users, and churned users — and then breaks each of those into their inputs — gives a diagnostic capability that the north star alone cannot provide. A drop in DAU without a metric tree requires days of analysis to diagnose. A drop in DAU with a metric tree can be diagnosed in minutes by checking which leaf node moved. Building the metric tree is the work that turns a north star metric from a number you watch into a number you can act on.