Not the smallest thing you can build. The smallest thing you can learn from.
The instinct when building something new is to keep adding features until it feels finished — but that risks spending months building toward a version of the product nobody actually wants, and only finding out at the end. An MVP inverts that: ship the minimum that lets real users interact with the core idea, and let their actual behaviour, not your assumptions, decide what gets built next. A 2012 academic treatment of the concept frames it specifically as a tool for reducing wasted experimentation in early-stage technology startups — building to learn, not building to launch.
This is exactly the discipline this essay admits to lacking: “too many times I kept finding myself trying to sort out small issues instead of building an overall MVP” — the failure isn’t a lack of skill, it’s skipping the discipline of shipping something minimal and testable before polishing.
Source: Moogk, D.R. (2012), “Minimum Viable Product and the Importance of Experimentation in Technology Startups,” Technology Innovation Management Review, March 2012. The term was popularised for a general audience by Eric Ries’s The Lean Startup (2011, Crown Business) — a practitioner book, not peer-reviewed itself.