Failure should not be big "all or nothing" events. A healthy culture should tolerate failed steps along the way to a successful project. A few of the early approaches or technologies should be high risk, but the learning and self-correcting of the organization should tolerate some failures and direction changes as healthy to improve the final product. One of the good aspects of a marketplace is the competition of ideas that should lead to survival of the fittest products.
Of course, if an organization does not have the stomach for risk or corrections, it should hang back and let others take the lead. Whether or not they can catch up is another issue.
Speaking of "failure", I'm surprised no one pointed out that a marathon is traditionally 26 miles and 385 yards, not the 37 miles you refer to .
Keep Up with TechRepublic