Minimum Viable Product
Build the smallest possible version that delivers core value to test assumptions.
Explanation
Eric Ries concept embraced by YC: an MVP isn't a smaller version of your final product—it's the smallest experiment that tests your core hypothesis about what users want. The goal is learning, not building. Many founders build far more than needed for initial validation, wasting time and resources.
Real-World Example
Dropbox's MVP was a simple video demonstrating file syncing, not actual software. Zappos started by posting shoe photos online and buying from stores when orders came in. Airbnb's MVP was air mattresses in the founders' apartment. Buffer started as a simple landing page with pricing—no actual product—to test demand.
How to Apply
Identify your core assumption about what users want. Design the smallest experiment to test this assumption. Build only what's needed for that test. If it's a feature, can you fake it with manual work? If it's a product, can you start with a landing page? Focus on learning speed, not building speed. Each MVP should test one key assumption.