Some authors claim many markets are winner takes all, where the dominant firm takes 75% or more of the profits, with the rest split between one or two followers

Some authors claim many markets are winner takes all, where the dominant firm takes 75% or more of the profits, with the rest split between one or two followers. So if you want to learn how to start one of those I think strategy is critical, not trying to retrofit one onto an existing product. It's like any architecture. Better to do the thinking before the building. But yes, it's far easier to learn anything when you're applying it too. Sure you could learn by trial and error but why not take advantage of the collective learning of an industry?