With all of the available buzzwords surrounding startup enthusiasts, one of the top phrases is MVP: Minimal Viable Product. Although I do believe in the concept, it has gotten so popular that people are quick to use it without genuinely understanding it. With no fault to Eric Reis, as he’s clearly explained the idea of MVP, people are just quick to dump a less-than half-ass product in the name of ‘MVP’ and then look bewildered when they don’t get proper support or traction.
Within the industry there’s conflicted views on the topic; Seth Godin isn’t a fan for many software startups. Then there’s Reid Hoffman’s #6 Rule which states that you should “Launch early enough that you are embarrassed by your first product release“. Unfortunately, many tech entrepreneurs are interpreting it as — release crappy, buggy, pre-alpha releases with the idea of getting great feedback from those sophisticated users.
Recently Robert Scoble posted about his irritation and why he’s criticizing start ups more critically lately. He points out his frustration with the term MVP, however if you read closely it’s really because too many people are using it as a crutch. There is a clear difference between a Proof of Concept, Minimal Viable Product and a Piece of Shit.
When you’re testing you’re hypothesis, or just flat out developing your application. Hopefully you have some type of plan. Taking that plan into account, this is where you can, and many people produce something that is a proof of concept to their idea. It can be fleshed out user interface design mockups, interactive mockups, non-functioning code (aka vapor-were) or even an pre-alpha. At this stage is where it is completely acceptable to have bugs, issues and non-functioning feature sets. It’s a proof of concept. It’s to help flush out the theories and see how people interact with it, what they think, etc.
The real problem arises when people try to pass off their PoC as their MVP. Don’t.
I believe (and hope) what Reid Hoffman meant was that, because you’re soooo passionate about your product and you know what you have planned — you’re initial release should be simplified to just the core (still functional, just simplified) so much that you’re uncomfortable because it’s so bare. It isn’t fully matured yet which is why you should be embarrassed to release that v1. Not because it has bugs, weird quirks and only 2 of the released 8 features really work.
You can still release a polished product as a MVP. Many startups do it. In essence that is the point, because how can you get genuine feedback on the concept when half of your feedback is related to bugs?
So yes, release soon and often, but only proper versions.
* update *
Great discussions going on via comments in Hacker News: http://news.ycombinator.com/item?id=3416911