As far as I understand it, MVP is all about determining whether there might be customers out there for you. (It's answering the question "Is anyone actually going to buy this thing?").

So staying small and out of the public spotlight to iterate on failure is valuable (and doesn't produce any negative PR).

If you have to go open beta to get that feedback, then do so. If you can get away with a private beta, do so.

Here's the better question: Can you set up a system so that you don't even need any sort of beta to tell?

Short answer is: It depends. Why would you want to release as a private beta?

You probably want users to be able to freely refer your product to friends. You probably want people to be able to discover, sign up, and try your product in one sitting. You probably don't want potential users to lose enthusiasm while they wait for a beta invite. You probably won't have crushing server traffic. If you do, you probably won't regret letting so many people use your product at once.

Thanks for the input. Agreed. I come from a design background so releasing something hacked together is a challenge for me. Furthermore, it really is a MVP-- there are definitely UX issues and functionality I would like to address/add, but resources are limited.

Probably depends on how many people or customers you have lined up for your private beta. If it's of sufficient size, a private beta can work just as well as a public launch.

Frankly, this is a sort of hack job @pud would be happy about it. Had an idea, busting it out in a few days and seeing what happens.

What could be cooler for a few days work to be out in open beta? You get feedback from real users (not friends/well wishers) and if it works well or sucks, you at least get honest feedback.

Agreed -- tangible honest feedback from real people is invaluable. Thanks for the feedback.