The term originally came from Eric Ries’s book “The Lean Startup”. The idea at the heart of the book is (roughly) that the job of a startup is to discover a viable business, and it should conduct this discovery as a series of hypothesis-driven experiments. One of the key principles is that an idea for a business contains any number of hidden assumptions, so a minimum viable product is the smallest, cheapest offering that will test the most basic assumptions.
The idea at the heart of the book is powerful and has caught on. I’m quite sure most people use the phrases “lean startup” and “MVP” without knowing their original meaning or even their genesis. In particular, MVP is one of those phrases for which it is easy to infer a similar-but-different meaning, and for that meaning to carry you a very long way in a conversation without the error being discovered.
I’ve been in conversations where MVP has been taken to mean any of “first release”, “prototype”, “a reduction on my original requirements”, “less than I would have liked” or “everything”. People too often think it’s just a modern digital phrase for a well-established concept, without realising it was intended to carry a specific and novel idea.
Personally I now try to tiptoe round the term and use words that are less open to abuse.