In the GenAI rush, are we forgetting PM 101?

🔥 Hot take: While everyone's scrambling to master the latest AI tech, the most successful products are still built on rock-solid Product Management fundamentals. 

Remember CRISP-DM Step 1? Business Understanding First!

Before I lose myself in the maze that is neural networks and transformer models, let's pause and remember what data science taught us: Business Understanding comes FIRST in the CRISP-DM process. Not the shiny tech. Not the coolest model. It’s the problem we're solving! 💡

(Side note: This was the catalyst that had me reflecting about this. I have championed this way of thinking as part of my operational excellence, ways of working, and Product Management background. My teams can attest how many times I’ve led with “What are we trying to achieve?”. So it truly was a wonderful “Aha!” moment when I got introduced to CRISP-DM by way of Duke University’s AI Product Management Course.)

AI Winners Who Got The Basics Right

🤖 OpenAI ChatGPT understanding the need for a versatile and accessible conversational AI.

🛡️ Anthropic Claude addressing the risk of AI generating harmful, biased, or non-compliant content.

💼 Microsoft 365 Copilot solving for productivity drain and workflow friction. 

All of the above prioritized making tools that are more capable, intuitive, and helpful in people’s daily lives over blinding, flashy “just because” features. If that isn’t foundational PM thinking, I don’t know what is.

Now don't get me wrong. Learn and build the new AI tech, yes! Truth be told I myself am catching up on these, getting me back to my Physics roots with all of the prompt engineering, algorithms, hyperparameter values, and loss functions. Beyond that, I’m looking forward to staying on top of:

• AI product metrics 📈

• Human-AI interaction patterns 🤝

• Ethics and bias considerations ⚖️

But we must all remember: Technology is the how, not the what or why.

A reminder of elements of our unshakeable PM foundation

🔍 Deep Human-Centered Research

• What jobs are our users hiring your product for? #JTBD

• What's their emotional journey, not just the functional one?

📋 Clear Problem Definition

• Can we articulate the problem in one sentence?

• How do we measure success beyond vanity metrics, and how do these ladder up to business objectives? #OKRs

🧪 Hypothesis-Driven Development

• What assumptions are we testing with each feature? #experimentation

• How do we measure effectiveness beyond "it works"?

What I can tell you is, for the talented, high-performing teams I've been part of, whenever we've stayed true to the above, we were able to create products that customers love and work for the business. 🎯

Discussion Starter:

If we stripped away all the AI buzzwords from our roadmaps and strategies, would our product still make sense to our users? Or have we become so enamored with the "how" that we've lost sight of the "why"? 🤔

#ProductManagement #GenAI #AIProductManagement 

Sources:

• Duke University AI Product Management Specialization

• OpenAI: Our Approach to Data and AI

• Anthropic: Our Approach to User Safety

• Microsoft: AI-powered success 

This post was crafted collaboratively with GenAI assistance. Here's to putting into practice what I preach about leveraging technology while maintaining human insight and strategic thinking. 🤖✨