Most early teams track either nothing or everything, and both fail the same way: nobody knows whether this week was better than last week. The fix is a single north star metric that captures the value you deliver, supported by a tiny cast of secondary numbers.
What Makes a Good North Star
A useful north star measures delivered value, not activity around it. For a hypothetical file-sharing tool, files actually shared per week beats accounts created, because sharing is the value and signups are only the doorway. Good north stars move when customers win, are hard to inflate with marketing stunts, and can be influenced by everyone on the team from support to engineering.
Why One and Not Five
Focus is the entire point. When a three-person team optimizes five numbers, it optimizes none of them,. One primary metric forces trade-offs into the open: a change that boosts signups but hurts weekly sharing is now visibly a bad trade rather than a mixed result everyone spins.
Changing It as You Grow
Your north star should evolve. Before launch it might be problem interviews completed; at early traction, weekly active usage of the core action; later, perhaps revenue retained and expanded. Change it deliberately at a stage transition; never drift between metrics whenever the news is bad.
Action Step
Write one sentence naming your product’s core value moment, then propose a north star metric that counts how often customers reach it. Test it against three questions: does it reflect delivered value, can the whole team affect it, and would gaming it be obvious. Adopt it for the next month.
Educational content only — not legal, financial, or individualized business advice, and no funding, hiring, or revenue outcome is promised. Every example is hypothetical; adapt the frameworks to your own situation.