# Product Analytics

> When to activate: product analytics, funnel analysis, cohort analysis, retention curves, DAU WAU MAU, activation metrics, event tracking, Mixpanel, Amplitude

- Skill: `mattakushi432/product-analytics` (Agent Skill)
- Install (CLI): `npx skillmds@latest add mattakushi432/product-analytics`
- Raw SKILL.md: https://api.skillmd.com/api/skills/mattakushi432/product-analytics/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Data & Analytics
- Author: Mattakushi432 (https://skillmd.com/u/mattakushi432)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/mattakushi432/product-analytics

---


# Product Analytics

## Event Taxonomy Design

### Naming Convention: Object-Action
```
Format: object_action

Examples:
  user_signed_up
  project_created
  file_uploaded
  subscription_upgraded
  report_exported
  onboarding_completed
```

### Event Properties (always include)
```json
{
  "event": "project_created",
  "user_id": "usr_123",
  "timestamp": "2024-01-15T10:30:00Z",
  "session_id": "sess_456",
  "platform": "web",
  "properties": {
    "project_type": "template",
    "team_size": 5,
    "plan": "pro"
  }
}
```

### Instrumentation Checklist
- [ ] Every screen/page view tracked
- [ ] All CTA clicks tracked
- [ ] Form submissions tracked (start + complete)
- [ ] Errors tracked with context
- [ ] Key milestones tracked (activation, upgrade, churn signal)
- [ ] Properties consistent across events (same field names)
- [ ] User identity linked (anonymous → identified on signup)

## Funnel Analysis

### Funnel Construction
1. Define the conversion goal (signup, activation, purchase)
2. Map required steps in order
3. Identify optional vs required steps
4. Set time window (complete within X days)
5. Measure conversion rate at each step

### Funnel Metrics
| Metric | Formula |
|--------|---------|
| Step conversion rate | Users completing step N / Users entering step N |
| Overall conversion | Users completing final step / Users entering step 1 |
| Drop-off rate | 1 - step conversion rate |
| Time to convert | Median time from step 1 to final step |

### Diagnosing Drop-offs
- Drop > 50% at a step → investigate with session recordings
- Compare drop-off by segment (device, plan, cohort)
- Run qualitative research at high-drop steps
- Test friction reduction (reduce fields, add progress indicator)

## Cohort Analysis

### Acquisition Cohort
Groups users by when they joined (week/month). Tracks how each cohort behaves over time.

```
Week 0: 1,000 users signed up
Week 1: 650 returned (65% W1 retention)
Week 2: 450 returned (45% W2 retention)
Week 4: 300 retained (30% W4 retention)
Week 8: 220 retained (22% W8 retention — stabilization point)
```

### Behavioral Cohort
Groups users by action taken (users who completed onboarding vs those who didn't).

Use behavioral cohorts to:
- Validate aha moment hypothesis
- Compare LTV of activated vs non-activated users
- Identify high-value behaviors to promote

## Retention Curves

### Benchmarks by Product Category
| Category | D1 | D7 | D30 |
|----------|----|----|-----|
| Social / consumer | 40% | 20% | 10% |
| SaaS (B2B) | 60% | 45% | 35% |
| Mobile games | 35% | 15% | 5% |
| E-commerce | 30% | 10% | 5% |
| Productivity tools | 50% | 35% | 25% |

### Reading the Curve
- **Smile curve**: Drops then flattens → healthy core audience
- **Declining curve**: Never flattens → product has no sticky value
- **High D1, steep drop**: Strong acquisition but weak activation

### Improving Retention
- D1 retention: Improve onboarding and time-to-value
- D7 retention: Establish habit loop, trigger return visit
- D30 retention: Deepen engagement, show accumulated value

## DAU / WAU / MAU

### Definitions
- **DAU**: Unique users with at least one qualifying event in a calendar day
- **WAU**: Unique users active in the past 7 days
- **MAU**: Unique users active in the past 28–30 days

### Engagement Ratios
| Ratio | Formula | Healthy range |
|-------|---------|---------------|
| DAU/MAU (stickiness) | DAU ÷ MAU | > 20% good, > 50% excellent |
| DAU/WAU | DAU ÷ WAU | Indicates daily vs weekly habit |
| WAU/MAU | WAU ÷ MAU | Weekly habit strength |

### Qualifying Event Definition
Do not count all events as active. Define a meaningful "active" event:
- Bad: any page view (inflated by bots/accidental visits)
- Good: core product action (created item, sent message, ran query)

## Activation Metrics

### Finding the Aha Moment
1. Define activation: what behavior correlates with long-term retention?
2. Segment retained vs churned users
3. Identify actions taken in first session that differ between groups
4. Test hypothesis: does promoting that action improve retention?

### Activation Metric Examples
| Product | Activation event |
|---------|-----------------|
| Slack | Sent 2,000 messages as a team |
| Dropbox | Uploaded 1 file on 2 devices |
| Twitter | Followed 30 accounts in first day |
| HubSpot | Connected email + created 1 contact |

### Time-to-Activate
Track: median time from signup to activation event.
Goal: reduce time-to-activate while maintaining quality of activation.

## Product Health Dashboard

### Key Metrics (top of dashboard)
| Metric | Timeframe | Alert threshold |
|--------|-----------|----------------|
| DAU | Daily | -10% WoW |
| New signups | Daily | -20% WoW |
| Activation rate | Weekly | -5pp MoM |
| D7 retention | Weekly cohort | -5pp MoM |
| Revenue (MRR/ARR) | Monthly | -5% MoM |
| NPS | Monthly | < 30 |

### Supporting Metrics
- Funnel conversion rates by step
- Feature adoption (% MAU using feature)
- Error rates on key flows
- Support ticket volume by category

## Mixpanel vs Amplitude vs PostHog

| | Mixpanel | Amplitude | PostHog |
|-|---------|-----------|---------|
| **Strengths** | Funnels, retention | Behavioral cohorts | Open source, session replay |
| **Pricing** | Usage-based | Usage-based | Free tier + usage |
| **Best for** | Growth teams | Enterprise | Startups, self-hosted |
| **SQL access** | Limited | Yes (paid) | Yes |