This is an excerpt from the book suggested by OverDriven some pages ago "Everything is obvious (*once you know the answer) : How common sense fails us". I took a look quickly and this excerpt retained my attention as there is some good analogy about cutting losses and pressing winners, maybe :
"Of all the prognosticators, forecasters, and fortune-tellers, few are at once more confident and yet less accountable than those in the business of predicting [...] trends.
[...]That these predictions are almost never checked for accuracy, that so many trends arrive unforeseen, and that the explanations given for them are only possible in hindsight, seems to have little effect on the breezy air of self-assurance that the arbiters [...] so often exude.
[...]Rather than trying to anticipate what shoppers will buy next season, Zara effectively acknowledges that it has no idea. Instead, it adopts what we might call a measure-and-react strategy.
First, it sends out agents to scour shopping malls, town centers, and other gathering places to observe what people are already wearing, thereby generating lots of ideas about what might work.
Second, drawing on these and other sources of inspiration, it produces an extraordinarily large portfolio of styles, fabrics, and colors—where each combination is initially made in only a small batch—and sends them out to stores, where it can then measure directly what is selling and what isn’t. And finally, it has a very flexible manufacturing and distribution operation that can react quickly to the information that is coming directly from stores, dropping those styles that aren’t selling (with relatively little left-over inventory) and scaling up those that are.
[...]traditional strategic planning invariably requires planners to make predictions about the future, leaving them vulnerable to inevitable errors—Mintzberg recommended that planners should rely less on making predictions about long-term strategic trends and more on reacting quickly to changes on the ground. Rather than attempting to anticipate correctly what will work in the future, that is, they should instead improve their ability to learn about what is working right now. Then, like Zara, they should react to it as rapidly as possible, dropping alternatives that are not working—no matter how promising they might have seemed in advance and diverting resources to those that are succeeding, or even developing new alternatives on the fly."
030985