Who This Helps
Product Managers who feel stuck choosing between five experiments and just want to know which one moves the needle. If you track 20 numbers but still can't decide, this is for you.
Mini Case
Meet Maya. She's a PM at a SaaS startup. Her team tracks 20 metrics, but every Monday she asks: "What should we test this week?" Last month, she ran three experiments. One improved activation by 12% in 7 days. The other two did nothing. Maya wasted two sprints on low-impact ideas. She needed a system to pick the winner before building.
Do This Now (5 Steps)
- Pick your North Star Metric. Choose one primary metric that defines success for your product. Keep it simple and clear. For Maya, it was "weekly active users completing the core action."
- Define three supporting metrics. These are leading indicators that feed your North Star. Maya picked "sign-up completion rate," "first action within 24 hours," and "7-day retention."
- Set realistic targets. For each supporting metric, write a target number for next week. Maya set a target of 40% for sign-up completion rate. This makes decisions concrete.
- Build a weekly scoreboard. Create a simple dashboard that shows your North Star, supporting metrics, and targets. Update it every Monday. No clutter. Just the numbers that matter.
- Use guardrails. Add one alert for each metric that triggers if it drops below a critical threshold. Maya set a guardrail: if 7-day retention falls below 25%, pause all experiments and investigate.
Avoid These Traps
- Tracking too many metrics. If you have more than five, you're not focusing. Cut ruthlessly.
- Setting vague targets. "Increase engagement" is not a target. Use a number like "increase weekly active users by 10%."
- Ignoring guardrails. Without alerts, you'll miss problems until it's too late.
- Changing metrics every week. Stick with your North Star for at least a month to see real trends.
- Forgetting to celebrate small wins. When you hit a target, take five minutes to acknowledge it. It keeps the team motivated.
Your Win by Friday
By Friday, you'll have a one-page scoreboard with your North Star metric, three supporting metrics with targets, and two guardrails. You'll know exactly which experiment to run next. No more guessing. Just calm, data-backed decisions.
And hey, you might even free up an hour for coffee. That's a win too.