- The Punk PM
- Posts
- The Punk PM #28
The Punk PM #28
Bias for Building
Hey there, punk!
I spent a couple of hours this week building a fantasy football scout app. Not because I had to. Not because it was on any roadmap. Just because I wanted to see if I could.
The tools were simple: ChatGPT, Magic Patterns, Lovable and Supabase. The goal was simpler: build something just for the fun of it, and to see how far I could execute on an idea.
Let's dig into why that matters more than ever.
Quote of the Week 🙊
Strategy is easy, execution is hard. Many companies have similar strategies, but what separates those who win from those who don't is execution.
— Deb Liu
Insight 🦉
Strategy, execution, analytics. That’s the PM toolkit in a nutshell.
Back in the ZIRP era, strategy was where you could stand out. When many PMs were stuck in feature factories, having a clear sense of why you were building something and how it connected to your business was a real advantage.
But in the AI era, execution is taking centre stage.
Prototyping tools have compressed the product feedback loop from months or weeks to days, sometimes even hours. PMs who can build (and build quickly) aren’t just shipping faster for the sake of it, they’re learning faster.
Because at its core, product management is all about learning. You can have the clearest strategy in the world, but it rarely survives first contact with your market. Mediocre strategy with great execution almost always beats great strategy with mediocre execution.
And by execution, I don’t mean rushing finished products out the door. I mean creating testable prototypes and tangible artefacts you can put in front of customers to get real feedback, sooner rather than later.
AI tools give us the chance to increase the pace of learning like never before. It’s worth embracing them. Just watch out for the trap of mistaking motion for progress. Most ideas won’t work out, and that’s okay. The quicker you learn that, the quicker you can move on to the next thing that might.
A bias for prototyping, testing, and experimenting helps you spot what isn’t working early, so you can find what will.
Speed of learning is your moat. Build to learn. Learn to build.
Action 🚀
This week, build something. Anything.
Pick a problem you've been thinking about and create a quick prototype. Use no-code tools, AI, or whatever gets you to a testable artefact fastest.
Don't aim for perfect, aim for done. You'll learn more from a rough prototype today than a perfect plan tomorrow.
Inspiration 💡
The Broken Mentality Around Software Development – Beekey Cheung challenges the factory-line thinking that plagues modern development. A sharp reminder that building software isn't about following processes, it's about creating something new. Read more
Inside Amsterdam's AI Welfare Experiment – Eileen Guo, Gabriel Geiger, and Justin-Casimir Braun unpack a fascinating case study in how not to automate decision-making. When Amsterdam tried to make welfare decisions more fair through AI, they learned some hard lessons about bias and execution. Read more
Fighting Product Enshittification – Janna Bastow breaks down why good products go bad and how to stop it. It's not just about features, it's about having the courage to keep building what matters, not what's easy. Read more
Signing Off ✍️
If this week's edition resonated, hit reply and let me know what you're building on the side. And if you've got a colleague who's stuck in thinking mode, maybe forward this their way.
Play it your way,
Toby