🧪MVP — The Most Misused Term in Tech
The Minimum Viable Product was never 'a stripped-down v1.' It was a test of a specific hypothesis.
Most teams use 'MVP' as a synonym for 'first version' or 'quick build.' This corrupts the original concept and produces bad outcomes. Understanding the real definition — a hypothesis test — is the difference between learning fast and shipping janky v1s.
An MVP (Eric Ries's original meaning) is the smallest thing you can build to test a specific hypothesis about customer value. It's a *learning* tool, not a *product* tool. The MVP is 'done' when you've learned whether the hypothesis is true, regardless of how polished the artifact is.
What MVP actually meant
Eric Ries, in The Lean Startup: an MVP is "that version of the product which enables a full turn of the Build-Measure-Learn loop with a minimum amount of effort and the least amount of development time."
The point is learning, not shipping. The MVP is an experiment with a hypothesis, a method, and a measured result.
How the term got corrupted
By 2014, "MVP" had come to mean "the first version of a product, stripped of nice-to-haves." It became the excuse for shipping low-quality builds with the promise of 'we'll improve it later.' That's not Lean Startup; that's just bad product.
Marty Cagan and others have written about this — the MVP, as popularly used, has done more harm than good.
The Aakash Gupta reframe
Don't ship an "embarrassing MVP" if you're past PMF. Once you have a product and customers, your reputation is the asset. A janky v1 of a new feature damages the perception of the whole product.
The right reframe: instead of MVP, think MVT (Minimum Viable Test). What's the smallest test you can run to validate your hypothesis? Often the answer isn't a built product at all — it's a landing page, a wizard of Oz, a concierge service, a prototype shown to 10 customers.
When MVPs (the real kind) are appropriate
- Pre-PMF. You don't know if anyone wants this. Build cheap, learn fast.
- New product category. No comparable benchmark; the only way to know is to try.
- High value-risk. You believe customers want X but no evidence yet. Build the smallest thing that proves or disproves it.
When MVPs are inappropriate
- Post-PMF, established brand. Your reputation is at stake. Ship complete, not minimum.
- Enterprise. Enterprise buyers won't try janky v1. Polish matters.
- Highly competitive category. You can't out-MVP a polished competitor.
What to do instead
For incremental work on a mature product: don't talk about MVPs. Talk about scope. Define what's in v1, in v1.1, in v2. Each is a real shipped artifact at a quality bar appropriate to your brand.
For new bets: instead of "let's build the MVP," ask "what's the cheapest test that would tell us if this is worth building?" Often the answer doesn't involve engineering at all.
Real-world examples
Before Dropbox built the file-sync system, Drew Houston made a 3-minute video showing how it would work. He posted it; the waitlist exploded from 5,000 to 75,000 overnight. That's a real MVP — a tiny artifact (a video) that tested a hypothesis (demand) without building the product.
Apple is famous for the opposite philosophy — they ship products only when polished. There's no MVP iPhone shipped in 2007 — the launch was a complete, high-quality device. The lesson: MVP isn't universal. Brand-driven companies have a higher minimum viable quality.
Go deeper — recommended reading
Interview questions (1)
Q1What's an MVP and how do you decide what's in scope?executionmid▼
An MVP in the original Eric Ries sense is the smallest artifact you can build to test a specific hypothesis. It's a learning tool, not a shipped product.
In practice, the term has been corrupted to mean 'janky first version.' I avoid that framing because it gives teams permission to ship low-quality work.
Instead I ask three questions:
- What hypothesis am I testing? Customer wants X? Channel Y will work? Build only what tests this.
- What's the cheapest test that would invalidate the hypothesis? Often it's not a built product — it's a landing page, prototype, or wizard of Oz.
- What's the quality bar appropriate to my brand and stage? Pre-PMF: scrappy is fine. Post-PMF: ship complete or don't ship.
The right scope falls out of these questions. The wrong scope comes from 'let's just MVP it,' which means nothing.