Chapter 2 · Product Analyst
2. Metrics, the heart of product analytics
~12 min read
If SQL is the analyst's hands, metrics are the analyst's judgment. Defining the right metric is the single most-tested and most-valuable skill in this role. This chapter gives you the frameworks senior analysts carry in their heads.
2.1 The North Star and the metric tree#
A North Star Metric is the one number that best captures the value your product delivers to users, and, when it grows, the business grows with it. Spotify's is time spent listening; Airbnb's is nights booked. Below it sits a metric tree: input metrics you can actually influence, each broken into leading indicators.
| Product type | Plausible North Star | Why it captures value |
|---|---|---|
| Music streaming | Time spent listening | Listening = the core value delivered |
| Marketplace (stays) | Nights booked | A booking is value for both sides |
| Messaging | Messages sent between people | Real communication, hard to fake |
| Collaboration tool | Weekly active teams | Team adoption = embedded value |
| Payments | Successful transactions | The job users hire it to do |
2.2 The AARRR framework#
Most product metrics fit into five stages of the user lifecycle, memorably called the 'pirate metrics' (AARRR). Use it to make sure your metric set is complete rather than random.
| Stage | Question | Example metrics |
|---|---|---|
| Acquisition | How do users find us? | Signups, CAC, traffic by channel |
| Activation | Do they reach first value? | Activation rate, time-to-value |
| Retention | Do they come back? | D1/D7/D30 retention, WAU/MAU |
| Revenue | Do they pay? | ARPU, conversion to paid, LTV |
| Referral | Do they invite others? | Viral coefficient, referrals/user |
2.3 The vocabulary you must know cold#
| Metric | Definition | Watch out for |
|---|---|---|
| DAU / WAU / MAU | Distinct active users per day/week/month | Define 'active' by a real action |
| Stickiness (DAU/MAU) | How many monthly users show up daily | Ratio, not a count |
| Activation rate | % reaching the first value moment | Pick the moment deliberately |
| Retention rate | % of a cohort still active after N days | Always cohort-based |
| Churn | % who stop using/paying in a period | The inverse of retention |
| Conversion rate | % completing a step | Always state numerator/denominator |
| ARPU / LTV | Revenue per user / lifetime value | LTV needs a retention assumption |
2.4 Vanity vs. actionable metrics#
| Vanity | Why it misleads | Actionable replacement |
|---|---|---|
| Total registered users | Never goes down; hides churn | Weekly active retained users |
| Total pageviews | Inflated by bounces & reloads | Sessions reaching a value moment |
| App downloads | Download ≠ use | Day-1 activation rate |
| Average revenue (mean) | One whale distorts it | Median + paying-user conversion |
2.5 Averages lie, segment first#
Two traps recur. The mean vs. median problem: a few heavy users drag the average far above the typical user. And Simpson's paradox: a trend in aggregate can reverse within every segment.
2.6 Business-model metrics every analyst should know#
| Model | Metrics that matter most | The number that kills you |
|---|---|---|
| Subscription / SaaS | MRR, churn, NRR, LTV:CAC | Churn outrunning acquisition |
| Marketplace | Liquidity, match rate, take rate | Supply/demand imbalance |
| E-commerce | AOV, repeat rate, contribution margin | CAC above gross margin |
| Ad-supported | DAU, sessions, time spent, eCPM | Engagement decline |
| Freemium | Free→paid conversion, activation | Free users who never convert |
LTV and CAC, the unit-economics pair#
CAC is what you spend to win a customer; LTV is the total margin that customer generates over their lifetime. A common rule: LTV:CAC ≥ 3:1 is healthy; below 1:1 burns money. LTV always rests on a retention assumption, which is why retention is the metric everything else depends on.
2.7 The HEART framework#
For UX-quality (not just growth), use HEART: Happiness (satisfaction), Engagement (depth), Adoption (new feature users), Retention, Task success (completion). AARRR for the growth funnel; HEART for the quality inside it.
2.8 A metric-definition workshop#
| Decision | Options | Why it matters |
|---|---|---|
| What counts as 'active'? | App open vs. core action | 'Open' inflates with notifications |
| Over what window? | Daily / weekly / monthly | Must match natural usage frequency |
| Unique or repeated? | Distinct users vs. sessions | Avoids double-counting heavy users |
| Which users? | All / new / paying | Changes whose behavior you're measuring |
Get the next chapter and weekly interview tips by email
One short email per week. Skim in a minute. Unsubscribe anytime.