Ask HN: How long was your web app beta period?

[shadowz]

For those of you that launched a web app, how long did it spend in private beta before official launch? It seems that the general consensus is to just put the product out there to test and iterate. What's your take?

[jacquesm]

15 minutes :)

We really jumped the gun and it is something I probably shouldn't have done but I never ever anticipated the kind of response it got.

By morning the server was a molten slag heap, I didn't even know because I was sleeping.

We mailed about 1,000 people that we were trying something new after building it, what we didn't anticipate was that they in turn mailed all their friends and so on.

A bit of performance testing before launch would have definitely helped.

The odd thing is, every time you do a lot of performance testing you end up with a dud :)

[coryl]

Private betas should only be used for products/businesses that really require private betas. I'm thinking that most startups don't really require private periods. We just go public and promote and see what happens.

[shadowz]

But isn't a focused private beta better than just a public launch?

[coryl]

Depends on your type of product and buzz you can generate imo. I mean, if your doing something new, and its practical that you have a limited amount of users, then a beta is great and make sense.

But most startups can't even generate the kind of buzz needed for a private beta, where it becomes worthwhile to build an invite/beta key system. If your product were to go public tomorrow, would your servers be killed in a traffic frenzy? If the answer is no, then you probably don't need a private beta.

[WiseWeasel]

Being able to control media coverage until you can be pretty sure the product works well is never a bad idea. A private beta allows you to do this.

[lleger]

None. We had a few close friends provide valuable feedback, and then we launched publicly.