The product decisions that matter
Most product decisions are reversible. Learning to identify the few that aren't changed how I approach building.
Early in my career, I treated every product decision like it was permanent. Database schema? Better get it perfect. API design? Can’t change it later. Feature scope? Locked in forever.
This made me slow and anxious. Everything felt high-stakes.
Most decisions are reversible
Here’s what I’ve learned: most decisions can be changed later. You can refactor code. You can migrate databases. You can deprecate APIs. You can remove features.
The cost of changing these things isn’t zero, but it’s usually lower than the cost of waiting for perfect information.
The few that aren’t
Some decisions genuinely are hard to reverse. These deserve more thought:
Pricing model. Moving from free to paid, or changing how you charge, affects user expectations in ways that are hard to undo.
Core data model. Not the schema details, but the fundamental entities and relationships. Getting this wrong creates friction forever.
Platform dependencies. Building on a platform means accepting their rules. Switching costs compound over time.
The practice
Now when I face a decision, I ask: “If we’re wrong, how hard is it to change?” If the answer is “pretty easy,” I move fast. If the answer is “very hard,” I slow down and think more.
Simple framework. Surprisingly effective.