This question originally appeared on CoFoundersLab: At what point do you sacrifice perfection in order to release a product?
Answer from Stevan Vigneaux, Director of Product Management and Marketing at Mimio, acquired by Boxlight
The MVP is crucial, and probably wrong because the team is likely still in the "This is what we want to build" phase even though everyone sat around the table and promised to only define the true and real "minimum." Feature creep doubtlessly happens, despite pledges.
There are different answers to this question for hardware and software products.
Hardware needs to work well enough and be reliable enough to do at least some significant portion of the intended task. Think of an amazing new car jack -- maybe your current version can only lift small cars, but not big pickup trucks. Okay, this is still potentially viable for a subset of the total market. But if your car jack can only lift a couple hundred pounds, then it isn't even ready for field deployment as a car jack. And if "every once in a while" it just collapses, then that's a death trap. Even if nobody gets hurt, people’s cars get broken because the jack collapses (and this is a death trap for your company). Not ready to release even in beta.
Software is generally different -- unless the software affects life and limb. After all, you should make sure that everyone who tests your product knows that "beta is not for sissies." But the metric for releasing is still much the same as for hardware -- it needs to work well enough and be reliable enough to do at least some significant portion of the intended task. But here "reliable enough" usually has a much lower bar. A video editor, for instance, might be cuts-only; it could be viable for testing the UI, processing speed, image quality, etc. If it crashes periodically, just make sure auto-save was the default mode; then they can reboot.
With those caveats, I'd release as soon as I have something that meets the core metric. Waiting for perfection lets "the other guys" get their product out the door ahead of you -- and first-mover advantage is not to be dissed. You may even find that the users seriously redefine your product. Which can cause you to say things like "But, that's not what you are supposed to do with my product" or "Whoa, that's not how this is supposed to be used." Don't say those things. Watch, listen, and discover, then think and think more ; the users might just be "righter" than you are. Or, they may have discovered product attributes and applications for which people are willing to pay and even evangelize.
However, there's the back side of that coin. We've all heard Henry Ford's comment "If I'd listened to customers I'd have spent my time breeding a faster horse." Customers generally lack vision. That's your job.
Still, release as soon as the product works well enough and is reliable enough to do at least some significant portion of the intended task.