The Lost Art of the Soft Launch

A lot the sentiments here echo a post from Eric Ries titled: "Lessons Learned: Don't Launch": http://www.startuplessonslearned.com/2009/03/dont-launch.htm...

In general, I think the lesson is that you should launch, refine your offering and target customer, and then do a big marketing launch.

Even restaurants do soft launches.

It's important, even as a "smoke test" (albeit in a bigger scale)

Slow rolling a launch reminds me of how the Facebook mafia tends to roll out their companies in measured and targeted ways, rather than trying to go viral all at once.

http://techcrunch.com/2011/02/13/inside-the-dna-of-the-faceb...

(scroll down to "controlled pacing")

I'm actually working on a new product (for a fairly small B2B niche) and planning to do a soft launch to work out all the kinks before I publicize it more widely. I'm thinking that'll allow me to work out any issues before I risk my reputation with people who might dismiss the product out of hand if they run into major issues. If you have more experience with it, does that make sense or should I forgo that first step?

Back when I did a lot more web stuff, soft launches were the default way we released features. But I wonder if soft launches are possible or even desirable in the mobile apps world where a lot of press is devoted to "new" (app store sections and even blogs with a recurring "new this week" sections).

Do any mobile startups do soft launches?