"Fail Fast, Fail Often" is really an indication that we need a Web Science

[amichail]

Why do all that experimentation yourself? Shouldn't it be done by web scientists once and for all? You could then use the results of their research as guidelines to determine the sort of app to build as well as its features/UI.

Also see:

http://www.cs.umd.edu/~ben/ShneidermanCACM6-2007.pdf

http://weblog.fortnow.com/2007/08/impact-of-facebook-platfor...

http://www.facebook.com/topic.php?uid=2205007948&topic=1...

[gibsonf1]

The biggest issue is something new. People aren't able to imagine the new big idea until they see it. When they see it, then they want it, but before that they had no idea they would want it. Case in point: the Iphone. Scientists are not the answer.

[nostrademons → gibsonf1]

Ugh, downvoted by mistake when I meant to upvote. Perhaps someone else can upvote to compensate? (And I really wish we could change our vote...it's pretty easy to slip...)

[amichail → gibsonf1]

You already rely on scientists in the implementation. So why not make use of scientific results in the specification as well?

Sure, the major idea can be new, but the way it is presented can make use of scientific results to increase the probability of success.

[nostrademons → amichail]

Many of us already do. That's why it's useful to keep up with recent results in psychology/advertising/marketing.

Problem is, one of those results is that consumers often decide whether to use a product based on an initial "gut feeling". That gut feeling is the combination of many factors, most of which are below the level of conscious perception. So science is effectively telling us that science cannot provide these answers...

[amichail → nostrademons]

The stock market is pretty unpredictable, yet people still try to identify trends.

I think it is too early to give up.

[nostrademons → amichail]

Yeah, and look where it's gotten us. ;-)

[gibsonf1 → amichail]

How would your scientific proposal work in the Iphone example?

[twism]

success is not a science

[amichail → twism]

Not completely, but partially it could be.