Ries’s definition is built around an operating condition rather than attributes: it contains neither size, age, industry, nor a garage. All three elements matter: an institution (not a product—a startup is an organization that must be built and managed), new products and services (not the optimization of what already exists), and extreme uncertainty (the defining condition itself).
The element “institution” yields an independent argument: a startup is an organization, a group of people, and an organization must be built and managed— management is necessary to a startup by definition, before any discussion of uncertainty. This institutional move independently supports A startup does not abolish management, but management for the predictable.
Consequences of defining it through a condition:
- “Entrepreneurs are everywhere”: intrapreneurship is legitimate, and “entrepreneur” is a job title in the organizational structure of companies that depend on innovation (principle 2).
- The boundary between “startup / mature company” is drawn not by revenue but by the share of activity conducted under uncertainty. The same organization contains both modes at once—and therefore needs both kinds of management simultaneously.
- Trend: “as the world becomes more uncertain, it gets harder and harder to predict the future”—the share of startup activity grows even in old organizations; a bridge to Organization 3.0 and the AI-native agenda.
Qualification: in the introduction, this expanded scope is asserted rather than proven—the only evidence there for applicability beyond software startups is the growth of the movement, that is, adoption rather than effectiveness (see Lean Startup is grounded in the genre it itself debunks).
[ET] Personal level: the definition also extends to personal activity—projects under conditions of high uncertainty are “startup-like” in type and require the same protocol of hypotheses and experiments; compare the concept of the Hypothesis-Driven Life project.
How it supports A startup does not abolish management, but management for the predictable
It supplies the condition from which the management dilemma follows logically: if a startup is defined by uncertainty, then management for the predictable is inapplicable to it by definition—and the choice is between chaos and experimental discipline.