~/learn/Product Metrics · Metric foundations
Vanity vs actionable metrics
Why cumulative totals look great and say little, and how ratios and cohorts tell the truth.
why this matters
Spotting a misleading chart is a quick way to show analytical maturity.
A vanity metric makes things look good but doesn't help you decide anything. The classic is total registered users: it can only go up, even if nobody uses the product anymore. Page views, app downloads and 'total messages ever sent' have the same problem.
Actionable metrics are tied to a decision and can go down. Weekly active users, conversion rate, retention by sign-up month and revenue per user all reveal whether things are getting better or worse.
From vanity to actionable
cumulative total → always rises → hides decline per-period count → shows growth or shrinkage ratio / rate → removes size effects cohort view → separates new vs old users
Cumulative users: 10k → 12k → 13k → 13.5k (looks like growth) New users/month: 2k → 1k → 0.5k (actually shrinking fast)
Interview trap
If an interviewer shows a cumulative chart and asks 'is the product growing?', point out that the slope (new additions per period) is what matters — a flattening cumulative curve means growth is slowing.
key takeaways
- Cumulative totals hide decline.
- Use per-period counts and rates.
- Cohorts separate new from old users.
check your understanding
Total registered users grew 5% this month. What can you conclude?