What is an MVP?
A Minimum Viable Product (MVP) is a version of a new product that allows a team to collect the maximum amount of validated learning about customers with the least amount of effort. It is not a smaller, cheaper version of your final product; it is a strategy and a process.
The core idea is to build a basic version of your product that targets a small, specific group of early adopters to test your most critical assumptions.
The Purpose of an MVP: Learning, Not Earning
The primary goal of an MVP is not to generate revenue or acquire a massive user base. The goal is to learn.
- Test Core Assumptions: Does your solution actually solve the problem for your target users?
- Validate Value Proposition: Is your unique value proposition compelling enough for people to use (or pay for) your product?
- Identify Key Features: Which features are essential, and which are just "nice-to-haves"?
- Reduce Risk: It prevents you from investing heavily in a product or feature set that nobody wants.
What an MVP is NOT
- A Crappy Product: "Minimum" does not mean low quality. An MVP should be well-designed, reliable, and usable for its limited feature set.
- A Beta Test: A beta test is typically about bug-finding on a near-complete product. An MVP is about testing fundamental business hypotheses.
- Just One Feature: An MVP must deliver a cohesive, valuable experience, even if it's very simple.
How to Define Your MVP
- Start with the Problem: Revisit your findings from Problem Validation and Customer Discovery. What is the single most important job your customer needs to get done?
- Identify the Core Workflow: Map out the essential steps a user would take to solve their problem using your product. What is the absolute minimum number of steps to deliver value?
- Prioritize Features Ruthlessly: Use a feature prioritization matrix (e.g., Value vs. Effort) to decide what to build. For an MVP, you should only build high-value, low-to-medium effort features.
- Choose Your MVP Type: An MVP doesn't even have to be code.
- Concierge MVP: You manually deliver the service to your first customers. This provides maximum learning with zero development.
- Wizard of Oz MVP: The user interacts with what looks like a fully automated system, but behind the scenes, humans are pulling all the levers. This is great for testing complex algorithms.
- Single-Feature MVP: A product that does one thing exceptionally well.
The Build-Measure-Learn Loop
The MVP is the cornerstone of Eric Ries's Build-Measure-Learn feedback loop.
- Build: Develop the smallest possible thing (the MVP) to test your hypothesis.
- Measure: Release it to early adopters and collect real-world data and feedback. Are users engaging? Are they completing the core workflow?
- Learn: Analyze the data and feedback to gain insights. Did you validate or invalidate your hypothesis?
Based on what you learn, you will either Persevere (continue with your current direction) or Pivot (make a fundamental change to your strategy). This loop repeats, with each cycle bringing your product closer to what the market truly wants.