| name | pm-case-study |
| description | Generate end-to-end PM case studies from real AI product launches, pivots, and decisions. Analyzes what happened, why, what the PM likely decided, trade-offs made, and lessons learned. |
| argument-hint | [product launch, feature, or company decision] |
PM Case Study Skill
Generate a detailed PM case study from a real AI product launch, pivot, or strategic decision — reconstructing the PM thinking behind it.
When to Use
- User asks "Write a case study on [AI product launch/decision]"
- User wants to understand PM decisions behind a real product
- User says
/pm-case-study followed by a topic
- Great for: ChatGPT launch, Claude's Constitutional AI, Gemini's multimodal strategy, GitHub Copilot pricing, Perplexity's search bet, Midjourney's Discord-first strategy, etc.
Framework: PM Case Study (8 Sections)
Section 1: Executive Summary
- What happened: One paragraph summary of the product decision/launch
- When: Timeline of key events
- Who: Key people and teams involved
- Outcome: How it played out (success, failure, mixed)
Section 2: Context & Background
- Company situation: Where was the company at this point? Stage, funding, competitive position.
- Market context: What was happening in the broader market?
- Technical context: What capabilities existed? What was newly possible?
- User context: What were users doing before this product? What pain existed?
Section 3: The Decision
- What was decided: Specific product/strategy decision
- Alternatives considered: What other paths were likely on the table?
- Key trade-offs: What did they give up by choosing this path?
- Stakeholder dynamics: Who likely championed this? Who likely opposed it?
Section 4: Execution Analysis
- Go-to-market strategy: How was it launched? Distribution channel?
- Phasing: Was it a big bang launch or phased rollout?
- Pricing: How was it priced? Why that model?
- Technical execution: What was the technical approach? Shortcuts taken?
Section 5: What Went Right
- Identify 3-5 specific decisions that contributed to success
- For each: What was the decision, why it mattered, what would have happened otherwise
- Be specific — reference actual features, timelines, or metrics where available
Section 6: What Went Wrong (or Could Have Been Better)
- Identify 2-3 mistakes, misses, or areas for improvement
- For each: What happened, what the impact was, what could have been done differently
- Be fair — hindsight bias is easy, focus on what was knowable at the time
Section 7: Metrics & Outcomes
- Growth metrics: Users, revenue, market share (use real numbers where available)
- Product metrics: Engagement, retention, satisfaction
- Strategic outcomes: Market position, competitive response, ecosystem effects
- Unexpected outcomes: Things that happened that nobody predicted
Section 8: Key Takeaways
Extract 3-5 lessons for product managers:
- Lesson: Clear statement of the principle
- Application: How to apply this in product sense/strategy decisions
- Example question: A product question where this lesson is directly relevant
Case Study Categories
Product Launches
- ChatGPT's launch (Nov 2022) — fastest growing consumer app ever
- Claude's positioning as the "safe" alternative
- Perplexity's answer engine vs. Google Search
- Midjourney's Discord-native strategy
- Cursor's bet on AI-native IDE
Strategic Pivots
- An AI lab's shift from nonprofit to capped-profit
- A safety lab's pivot from pure research to product company
- A big tech company's emergency response to ChatGPT
- An open-source LLM strategy from a major tech company
Feature Decisions
- ChatGPT Plugins → GPTs → the pivot to actions/agents
- GitHub Copilot's pricing model ($10/month individual)
- Claude's Artifacts feature
- Gemini's multimodal-first approach
- NotebookLM's audio overview feature
Pricing & Business Model
- LLM API pricing evolution (the race to the bottom)
- ChatGPT Plus ($20/month) → Team → Enterprise tiers
- The free tier strategy across AI companies
- Usage-based vs. seat-based pricing in AI
Output Format
Write as a business school case study — structured, analytical, and with clear takeaways. Use real data where available, clearly mark estimates or speculation. Aim for ~2500 words.
Research-First Workflow (CRITICAL)
This skill requires real data:
- Research extensively — Do 10-15 web searches for: launch details, user growth data, pricing history, company blog posts, founder interviews, analyst reports, and competitor responses.
- Cite everything — Include
[linked source](url) inline for all factual claims.
- Date awareness — Note what was known at the time of the decision vs. what we know now.
- Display the complete case study.
What Good Looks Like
- Demonstrates deep knowledge of the AI product landscape
- Shows you can analyze real product decisions with nuance
- Provides concrete examples and data points for product discussions
- Builds pattern recognition across multiple product launches
- Reveals your product judgment when you evaluate decisions
