FAQ Off

Venture funding and the companies it produced, from Fairchild in 1957 to the filings of the present decade.

2001Vocabulary · File 4.01

Frank Robinson Coined It; Eric Ries Made It Mandatory

The term 'minimum viable product' has two authors, two definitions, and one decade between them.

A workbench with a rough prototype half assembled, cardboard and tape, tools around it
2011 — Eric Ries publishes The Lean Startup; the term enters mainstream startup vocabulary with a revised definition

01 / VocabularyTwo Moments, One Term

Frank Robinson, co-founder of the product-management consultancy SyncDev, introduced the phrase "minimum viable product" in 2001. His formulation was commercial in orientation: the MVP was the version of a product that maximised return on risk for both the vendor and the customer — sized precisely so that the developer could ship without waste and the buyer could adopt without hesitation. Robinson's concept was a scoping tool, concerned with matching product scope to market readiness.

A decade later, Eric Ries gave the term its current meaning. In The Lean Startup (2011), Ries defined the MVP as the smallest release capable of completing a cycle of validated learning — that is, the fastest way to test a founding assumption against real users and discard it or build on it. Where Robinson's MVP minimised commercial risk at a single point of sale, Ries's MVP was an instrument of continuous experimentation inside a feedback loop he called Build–Measure–Learn.

A whiteboard covered in a handwritten cap table — percentage columns, founder rows, round labels — in a plain office
Robinson (2001) — the version of a product sized to maximise return on risk for vendor and customer simultaneously; a scoping and commercial decision
The printed cover of an S-1 registration statement lying on a wooden desk, corner turned, grain visible at extreme close focus
2001 — Frank Robinson of SyncDev introduces "minimum viable product" in a product-management contextPhoto: RDNE Stock project / Pexels

The definitional gap matters. Robinson's version assumes the product will ship and be sold; the question is how much to include. Ries's version is agnostic about whether the first iteration is a product at all — his canonical examples include a hand-edited concierge service or a video demonstrating a feature that does not yet exist. The MVP, under Ries, need not be viable in any commercial sense; it need only generate data.

The Lean Startup sold millions of copies and the term moved with it, embedding itself in Y Combinator batch cycles, accelerator curricula worldwide, and corporate innovation programmes. The Robinson origin was largely forgotten; most practitioners who use "MVP" daily are working from Ries's definition without knowing another existed. By the mid-2010s the phrase had grown vague enough that its utility as a precise instruction — which was the only thing either author intended — was routinely questioned by practitioners on both sides of the original divide.