The wrong question
Most founders scope their MVP by asking 'what's the smallest version of my product?' That question produces a stripped-down product, not a validated one.
The right question
The better question is: what is the riskiest assumption behind this business, and what's the smallest thing we can build to test it? Sometimes that's a full product slice. Sometimes it's a single workflow with everything else handled manually behind the scenes.
Architecture still matters
A lean MVP scope doesn't mean throwaway code. Early architecture decisions — data model, auth, core integrations — are expensive to unwind later, so they deserve real design thinking even when the feature set is deliberately small.
The takeaway
Scope your MVP around the assumption you're most afraid is wrong, not around a feature checklist. That's what makes it a minimum viable product instead of just a minimum product.
