News for hackers. Stuff that matters.
Re:Death to the Cult of Stealth Startups by dsirijus on Friday May 24, 2013 @11:58AM (#5763677)
Example of neccesary noise would be me trying to support clients who don't have a clue what they're doing with the product (and it's their stupidity to blame).

As for unneccesary noise example, that would be feedback of people whose feedback I do not value.

And as for edited and added last question - it depends. :)

Re:Death to the Cult of Stealth Startups by blacktar on Friday May 24, 2013 @12:16PM (#5763783)
When you say"support clients", I assume people already paying for what your startup is offering? For clarity, I'm explicitly talking about sharing for pre-product-market fit, pre-validated business model startups. (edit: sloppy fingers)
Re:Death to the Cult of Stealth Startups by dsirijus on Friday May 24, 2013 @12:22PM (#5763823)
Well, with your openess, you're out of the gate already, so "pre-" prefix isn't in the same sense as for stealth startup.

Dilluting my karma like I just don't care! :)

Re:Death to the Cult of Stealth Startups by blacktar on Friday May 24, 2013 @12:36PM (#5763917)
What do you mean? Could you elaborate how my startup is now in fact different than one hour ago before the post?
Re:Death to the Cult of Stealth Startups by blacktar on Friday May 24, 2013 @12:10PM (#5763754)
How do you know to objectively value feedback sources? Why and how does it depend?
Re:Death to the Cult of Stealth Startups by dsirijus on Friday May 24, 2013 @12:19PM (#5763804)
I usually bundle feedback in two groups - first one is my target demographic, the second one is professional. First one is pretty straightforward to parse, the second one is a kind of a double-edged sword.

Whatever your methodology is, at some point you can't parse more feedback, and it's then noise, whatever its contribution is.

Re:Death to the Cult of Stealth Startups by blacktar on Friday May 24, 2013 @12:27PM (#5763856)
I see it this way: The higher frequency of exposure, the larger the sampling pool, the more noise AND the more opportunity to spot a recurring signal in that noise. If you keep your sampling pool low, how will you separate noise from signal?