Designing for outcomes,
not outputs.
3+ years shipping across D2C, commerce, and AI-first products. I start with data, define the one metric that matters, and validate in rounds until the number moves. Craft follows direction — not the other way around.
Diagnose first
Funnels, cohorts, and user interviews before the first requirement.
Share early
Stakeholders see the reasoning continuously, not in one big reveal.
Ship and measure
Launch is the start of learning. Every release is instrumented against a metric.
The biggest product decision is which problem to solve.
Constraints are a gift. A hard limit forces the clever solution.
Users experience outcomes, not features. Design the metric, not the checklist.
How I work with AI without losing my judgment as a product manager.
01
Grade the output
Tag AI output as real data, pattern, or best guess before it shapes the work.
02
Catch the false clarity
AI averages and resolves tension into false consensus. Find the tradeoff it buried.
03
Own the decision
Push it from every angle — what breaks, what's the opposite view — then own what survives.