Back to PM Skills
Metrics & Experiments

Metrics Framework Builder

Build comprehensive metrics frameworks using AARRR pirate metrics or Aakash Gupta's input/output methodology.

# Drop into ~/.claude/skills/metrics-framework/
curl -L https://github.com/sunnyyang-hicks/pm-skills-for-claude/raw/main/metrics-framework/SKILL.md \
  -o ~/.claude/skills/metrics-framework/SKILL.md

Overview

You build metrics frameworks that connect daily product decisions to business outcomes. The goal: any PM on the team can look at the dashboard and know whether the product is healthy, what's broken, and where to dig deeper.

Before You Start

Ask the user:

  1. Product type — B2B SaaS, marketplace, consumer app, internal tool, API product.
  2. Business model — Subscription, transaction-based, ad-supported, usage-based.
  3. Stage — Pre-PMF, growth, scale, mature. This changes which metrics matter.
  4. Current state — What do you measure today? What's instrumented?
  5. Key decisions — What decisions should these metrics inform?

Framework Options

Option A: AARRR (Pirate Metrics)

Best for: Consumer products, PLG SaaS, marketplace businesses.

## Metrics Framework — AARRR Model

### Acquisition: How do users find us?
| Metric | Definition | Current | Target | Source |
|--------|-----------|---------|--------|--------|
| New signups/week | Unique accounts created | [X] | [Y] | [tool] |
| Signup conversion rate | Visitors → signups | [X%] | [Y%] | [tool] |
| CAC by channel | Cost per acquired user per channel | [$X] | [$Y] | [tool] |
| [channel-specific metrics] | | | | |

### Activation: Do they experience the core value?
| Metric | Definition | Current | Target | Source |
|--------|-----------|---------|--------|--------|
| Activation rate | % signups reaching "aha moment" | [X%] | [Y%] | [tool] |
| Time to value | Median time from signup to aha | [X days] | [Y days] | [tool] |
| Onboarding completion | % completing setup flow | [X%] | [Y%] | [tool] |

**Define your "aha moment":** [The specific action that correlates with long-term retention.
This is the single most important metric definition in the entire framework.]

### Retention: Do they come back?
| Metric | Definition | Current | Target | Source |
|--------|-----------|---------|--------|--------|
| D1/D7/D30 retention | % users returning after N days | [X%] | [Y%] | [tool] |
| Weekly active rate | WAU/MAU | [X%] | [Y%] | [tool] |
| Churn rate | % users lost per period | [X%] | [Y%] | [tool] |
| Feature retention | % using key feature repeatedly | [X%] | [Y%] | [tool] |

### Revenue: Do they pay?
| Metric | Definition | Current | Target | Source |
|--------|-----------|---------|--------|--------|
| MRR/ARR | Monthly/annual recurring revenue | [$X] | [$Y] | [tool] |
| ARPU | Avg revenue per user | [$X] | [$Y] | [tool] |
| Conversion rate | Free → paid | [X%] | [Y%] | [tool] |
| Expansion revenue | Upgrades + add-ons | [$X] | [$Y] | [tool] |
| LTV | Lifetime value | [$X] | [$Y] | [tool] |

### Referral: Do they tell others?
| Metric | Definition | Current | Target | Source |
|--------|-----------|---------|--------|--------|
| Viral coefficient | Invites sent × acceptance rate | [X] | [Y] | [tool] |
| NPS | Net Promoter Score | [X] | [Y] | [tool] |
| Organic share rate | % users sharing/inviting | [X%] | [Y%] | [tool] |

Option B: Input/Output Model (Aakash Gupta)

Best for: When you need to connect team actions to business results.

## Metrics Framework — Input/Output Model

### Output Metrics (Business Results — Lagging)
| Output Metric | Definition | Current | Target | Owner |
|--------------|-----------|---------|--------|-------|
| [metric] | [what it measures] | [X] | [Y] | [team] |

### Input Metrics (Team Actions — Leading)
For each output metric, identify the inputs that drive it:

**Output: [Revenue/Retention/Growth target]**
| Input Metric | Hypothesis | Current | Target | Lever |
|-------------|-----------|---------|--------|-------|
| [metric] | [why this input drives the output] | [X] | [Y] | [what team can do] |

### Causal Chain
[Output] ← [Input 1] + [Input 2] + [Input 3]
Show the math: "If we move Input 1 from X to Y, our model says Output moves by Z."

Implementation Checklist

After defining the framework:

  • [ ] Every metric has a precise definition (no ambiguity about how it's calculated)
  • [ ] Every metric has a data source identified
  • [ ] Every metric has a baseline measurement
  • [ ] Targets are grounded in historical data or benchmarks, not wishes
  • [ ] Counter-metrics are defined (what NOT to sacrifice for each target)
  • [ ] Dashboard is designed with metrics grouped logically
  • [ ] Review cadence is set (daily glance, weekly review, monthly deep dive)
  • [ ] Alerting thresholds are defined for critical metrics

Output

Save as METRICS-FRAMEWORK-[product-name].md.