Startup Metrics
What to measure, when, and why. Dave McClure’s AARRR framework (“Pirate Metrics”) provides the canonical structure. The core principle: measure what matters, ignore vanity metrics, and use data to drive decisions.
The AARRR Framework (Pirate Metrics)
Dave McClure (500 Startups, 2007) broke the customer lifecycle into five stages:
| Stage | Question | Key Metrics |
|---|---|---|
| Acquisition | How do users find us? | Traffic sources, signups, CAC by channel |
| Activation | Do they have a good first experience? | Onboarding completion, time-to-value, “aha moment” rate |
| Retention | Do they come back? | DAU/MAU, churn rate, cohort retention curves |
| Referral | Do they tell others? | NPS, viral coefficient, referral rate |
| Revenue | Do they pay? | MRR/ARR, ARPU, LTV:CAC ratio |
How to Use AARRR
- Map your funnel: Identify what each stage looks like for your product
- Measure each stage: Set up tracking for the key metrics
- Find the bottleneck: Where is the biggest drop-off? That’s your priority.
- Optimize the bottleneck: Focus all effort on improving that one stage
- Move to the next bottleneck: Once improved, the constraint shifts elsewhere
Common Mistake: Optimizing the Wrong Stage
Most startups over-invest in Acquisition (marketing) when their real problem is Activation (users don’t get value) or Retention (users don’t come back). Pouring users into a leaky bucket is waste.
The Growth Loops Critique of AARRR
Brian Balfour (Reforge) argues AARRR is fundamentally limited as a growth model: it explains individual steps but cannot answer “how does your product grow?” Funnels are linear with no concept of reinvesting output as input, and they create organizational silos. The replacement framework is growth-loops — closed systems where one cohort of users creates conditions for the next, producing compounding rather than linear growth. AARRR remains useful for measuring steps within a loop, but the loop itself is the unit of growth strategy.
Vanity Metrics vs Actionable Metrics
| Vanity (Avoid) | Actionable (Track) |
|---|---|
| Total signups | Active users (DAU/MAU) |
| Page views | Conversion rate |
| App downloads | Retention by cohort |
| Social media followers | Revenue per user |
| Press mentions | NPS / referral rate |
Altman: “Avoid vanity metrics — focus on retention alongside acquisition.”
The One Metric That Matters (OMTM)
At any given time, pick one metric as your north star:
- Pre-PMF: activation rate or retention (are people getting value?)
- Post-PMF: revenue growth rate or unit-economics (can you scale profitably?)
- Scaling: efficiency metrics (CAC payback, gross margin)
Post it visibly. Altman: Airbnb’s founders posted growth graphs everywhere.
Rabois’ Paired Metrics
From operations: never measure one metric alone — it invites gaming. Pair complementary metrics:
- Fraud rate + false positive rate (forces innovation, not crude blocking)
- Growth rate + churn rate (prevents “growth” that’s just a leaky bucket)
- Revenue + customer satisfaction (prevents short-term extraction)
Stage-Appropriate Metrics
| Company Stage | Primary Metrics | Why |
|---|---|---|
| Pre-PMF | Qualitative feedback, Sean Ellis score | You’re searching, not optimizing |
| Post-PMF | Retention, revenue, growth rate | Validate the model works |
| Scaling | CAC, LTV, payback, gross margin | Prove unit-economics |
| Growth | Market share, efficiency, profitability | Sustain and defend |
See Also
Backlinks
- ai-evals
- balfour-growth-loops
- case-study-wework
- cohen-max-mrr
- distribution
- focus
- founder-faq
- founder-faq-slides
- founders-operating-system
- founders-operating-system-slides
- glossary-of-frameworks
- growth-loops
- onboarding
- pitching
- product-development
- product-led-growth
- retention-and-churn
- second-time-founders
- shankar-husain-ai-evals
- start-here
- the-saas-playbook
- walling-saas-playbook