Metrics and experimentation
Product metrics and the north star
Choosing a north star that reflects customer value, decomposing it into input metrics teams own, and pairing it with guardrails. Start here for the map.
A product team without a shared metric argues about features. A team with a bad metric ships growth that looks good on a slide and quietly destroys the business. The job of product metrics is to make customer value visible enough that engineers, PMs and marketers pull in the same direction — and to make damage visible before you celebrate a false win.
The centerpiece is the north star metric: one number that best captures the value your product delivers, that tends to lead revenue and that teams can actually move. Around it sit input metrics (levers you control), guardrails (things you must not break) and a ruthless allergy to vanity.
As an engineer, you will be asked to "add metrics" or "instrument the funnel." Push for definitions: event name, properties, identity and which dashboard number this event feeds. A north star with no event taxonomy is a wish. The instrumentation layer is product analytics.
Concepts in this section
Read them in order. Each one builds on the one before.
| Step | Concept | What it answers |
|---|---|---|
| 1 | The north star metric | Which one number captures customer value and can teams move it? |
| 2 | Metric trees | Which input stalled, and which lever is cheaper? |
| 3 | Guardrails and vanity metrics | What must we not break, and which numbers only feel good? |
| 4 | AARRR and leading metrics | Where on the journey is the tree broken, and can we steer this week? |
When a PM proposes a new input, ask three questions: Can we move it in six weeks with product work? If it doubles and the north star does not move, what did we learn? What guardrail stops us from gaming it? If they cannot answer, the metric is not ready.
In interviews, structure the answer: customer value → candidate metric → why it leads revenue → 3 inputs → 2 guardrails. Mention one way it could be gamed.
Common mistakes
- Copied north stars. Copy the criteria (value, leading, actionable), not Airbnb's noun.
- Local optima. Acquisition floods low-intent signups; activation looks worse while growth celebrates CAC. Review the tree.
- Output theater. Story points on the product scoreboard. Keep delivery metrics on the eng ops dashboard.
- Under-instrumented ambition. A Looker tile with no events behind it.
Key takeaways
- A north star quantifies customer value, leads business results and is actionable.
- Decompose it into a tree of inputs with owners; pair it with 2–4 guardrails.
- Prefer actionable outcome counts over vanity; AARRR maps the journey, it does not replace the star.
- Change the north star when strategy changes — not when the chart is inconvenient.
Further learning
Amplitude's playbook is the clearest practical treatment of north star + inputs.
The North Star Playbook (Amplitude)Put the North Star Framework to work for your product. Get expert tips on defining and integrating your North Star Metric into your development processes.Amplitude
16 Startup Metrics (a16z)We have the privilege of meeting with thousands of entrepreneurs every year, and in the course of those discussions are presented with all kinds of numbers, measures, and metrics that illustrate the promise and health…Andreessen Horowitz
Ronny Kohavi on why most ideas fail in controlled tests:
Related
Product analytics
Designing event tracking plans, reading funnels and cohorts, writing the SQL that answers product questions, and avoiding mix-shift illusions. Start here for the map.
A/B testing and experimentation
Designing randomized experiments, sizing samples with real arithmetic, reading lift and confidence intervals, and shipping only what the data actually supports.
SaaS metrics: ARR, churn and net retention
ARR and the bridge, bookings vs revenue, logo vs revenue churn, GRR, NRR, and the quick ratio.