Writing
Ongoing thoughts on AI product leadership, experimentation culture, personalization systems, and what it means to build platforms that last.
8 pieces
There's a distinction that matters enormously in product strategy: building an AI feature versus building an AI decisioning system. One adds a capability. The other changes who or what is making choices — and that changes everything downstream.
Organizations often try to fix experimentation quality by adding more documentation, more approval gates, more templates. That's the wrong instinct. Culture problems need cultural solutions — which means changing what gets rewarded and what gets scrutinized.
The trap most personalization teams fall into is optimizing for engagement signals that reflect past behavior rather than present intent. You end up building a mirror instead of a window.
The role of a product leader hasn't changed at its core — it's still about judgment, alignment, and customer empathy. But the context has shifted dramatically. AI changes the pace of iteration, the nature of failure modes, and the kinds of expertise that need to be in the room.
Building platforms requires a different kind of product thinking than building features. You're designing for builders, not users. The decisions you make get amplified across every team that depends on you — which means the cost of a bad platform decision compounds in ways feature decisions don't.
Every experienced product leader has a story about a metric that looked great and meant nothing. Or worse — pointed in the wrong direction. The discipline of choosing what to measure is as important as the discipline of measuring well.
Many AI products are designed to impress at launch and forgotten six months later. The ones that last are built around feedback loops — systems that improve with use and deteriorate predictably when they don't get the signal they need.
The gap between executive excitement about AI and executive understanding of AI investment is where most ambitious AI roadmaps go to die. Bridging that gap is a product leadership skill — and it's learnable.
More writing coming soon. Follow on LinkedIn for the latest.