Axoniq Conference 2025 Day 2: Event Sourcing from a PM Perspective - Learning & Advocacy

🎯 Speaker:

Michael Mathieu - Product Manager (Manufacturing → Wayfair → Product)

💡 Key Innovation: 4 User Story Types

Based on event sourcing concepts:

1️⃣ Command Story

• As a user, I want to DO SOMETHING so that I can REALIZE VALUE

• Do something = Command

• Realize value = Event

2️⃣ Query Story

• As a user, I want to KNOW SOMETHING so that I can MAKE A DECISION

• Know something = Query

• Make decision = Outcome

3️⃣ Consequence Story

• As somebody, I want a command/event to HAVE A CONSEQUENCE so that I can MOVE THE PLOT FORWARD

• Background processing, validation, automation

4️⃣ Efficiency Story

• As somebody, I want the above to happen MORE EFFICIENTLY so I can REALIZE MORE VALUE FASTER

• Optimization, performance improvements

🎯 Selling Event Sourcing:

Don't Sell:

• Technical architecture

• CQRS, DDD, Event Sourcing terminology

• Scary acronyms to business

Do Sell:

• Auditability capabilities

• Rollback functionality

• Business outcomes

• Problem-solving approach

💼 PM Responsibilities:

• ✅ Educate yourself first (read, tinker, ask engineers)

• ✅ Convince business stakeholders (your job, not engineers')

• ✅ Engineers convince architects (their job)

• ✅ Use business-friendly analogies (bank transactions, life events)

• ✅ Focus on capabilities, not architecture

• ✅ Invest political capital to get approval

🏗️ Building with AI:

• Used Claude to build UNESCO site tracker

• Learned TypeScript, then Axon Framework

• Lovable.ai built front-end from Swagger JSON

• Claude generated 120 user stories from existing code

• AI as learning tool, not replacement

🔑 Key Analogies:

Life is Event-Sourced:

• Bank account = series of transactions

• Growth = books you read + people you meet

• Stroop waffle lifecycle: born → first waffle → Amsterdam → death

• Eventually consistent = paperwork in Germany

Manufacturing Connections:

• IBA Analyzer: production events in steel mills

• Value stream mapping = event modeling

• Kanban boards (1970s) = modern product management

• Everything stamped with timestamps

📋 Product Triad:

Business ← → Product ← → Engineering

          ↓              ↓

       Problems      Solutions

PM owns the business relationship and must sell technical decisions upstream.

🚫 Common Mistakes:

• ❌ Letting business dictate solutions (database of things)

• ❌ Not educating PMs about architecture

• ❌ Engineers trying to sell to business directly

• ❌ Using technical jargon with stakeholders

• ❌ Ignoring political capital requirements

✅ Success Factors:

• Strong interpersonal relationships

• Understanding the actual problem

• "What problem are you trying to solve?" (repeat 7x)

• Pragmatism about highest-paid opinions

• Right timing (before code in production = one-way door)