Beware the tortoise,
its speed has boundless limits,
amidst snails and rocks
This likely speaks more of my own idiosyncracies than anything else, but I've always been more fascinated about how people fail than how they succeed. Without failure and without anything to overcome, how does anyone expect to improve? More importantly, how do you know when you've failed? When your transgressions of curiosity, even with the most creative of bookkeeping tactics, shifts the balance firmly into the red? You often read about how the most successful entrepreneurs exhibit tremendous bravery in "disrupting" industries by approaching problems in ways that more traditional mindsets would discourage. You don't often read about the other 99% of entrepreneurs who hung onto their ideas just a tad too long, situations where the best case scenario is to start from square one, if only they're so lucky.
Outside of established companies, I think the benchmark(s) for failure and success becomes much hazier. Maybe that's the real appeal of research and startups. We have some flexibility in redefining the definition of success while establishing new fields and approaches. In our lab, we often joke about metrics and frameworks in the context of academic papers. Those who propose the most reasonable new metrics and frameworks often get the bulk of future citations, even if there's little practical or even experimental applications. The slow cycle of research, without the benefit of a market for even the most basic of beta-testing, means that entire academic careers could be built on the flimsiest of premises. Likewise, it seems that some startups can survive on hype and promises alone.
I guess one counterargument would be that any new approach requires time to validate. Who are we to say that past failures would have stayed failures if they were given the opportunity to adjust? Maybe Webvan would still be around today if their investors had the cajones to provide a few extra tens of millions (I kid...I think).
The fascinating thing to me is the set of decision points that forces someone to abandon their approach. Failure (to me) has always seemed easier to quantify and recognize than success, but always far more difficult to accept. In my opinion, the crazy thing about trying truly new approaches is that you're effectively saying that the status quo is wrong without knowing whether or not you're right. Creativity makes you either a laughingstock or a genius, so how do you know when you've gone off the deep end?
Like I said, without the corrective effect of actual markets, research seems to exemplify the problems associated with not recognizing failed approaches to nowhere. It's not inconceivable (and perhaps it's common) for students to spend an entire PhD on a solution to a problem that turns out to be completely invalid. As you can imagine, trying to build a career off of a research topic with no future is as difficult as trying to build a business off of a product with no demand.
The nerve-wracking frustration that I'm faced with daily is that you may never truly know until everything's said and done, so how do you decide? There's a certain amount of bull-headed arrogance and reckless bravery (or stupidity) to break away from the status quo. "Fail fast and fail often" is probably the easy solution, but in a reality where most lack the academic or financial capital to do so, what is one to do?