Minimum Viable Products vs Minimum Viable Tests

There has been a lot of confusion with the terms minimum viable product and what I consider minimum viable tests.

The minimum viable product is a term that is based off of the “minimum feature set” that Steve Blank often refers to.  Although Steve refers to these terms interchangeably, the minimum feature set is the minimum features of a product or service that will make early adopters interested in the product.  The minimum feature set is just that, a list of features.  Once you build an actual product or prototype that early evangelists can use, then you have a minimum viable product.  A minimum viable product is the product you build using the minimum feature set.

So the minimum viable product is an actual product.  It’s something viable that at least one early evangelist will use.  The minimum viable product comes after you “get out of the building”.  So what’s the minimum viable test?  The minimum viable encompasses strategies when interacting with potential users that help you determine the minimum feature set.  When you interview a prospect to find out their typical day, that is a minimally viable test.  When you draw sketches and show them to users in a customer development interview, that’s a minimally viable test.  When you create fake landing pages to see if a user will click the “buy now button”, that’s a minimally viable test.  It’s called a minimum viable test because you’re using the minimum amount of effort to validate or invalidate your hypotheses. 

Dropbox and their “minimum viable product”

Most people refer to Dropbox’s original video as their minimum viable product.  In reality, that video was an MVT (minimum viable test).  Their actual Minimum Viable Product wasn’t released until about a year later after they went through the ycombinator program.  They realized their minimum feature set soon after the video launched.  The video was a great help in determining that their hypotheses were correct.  To build the software took a much longer time than it took to validate the hypothesis.