I think this is a wrong question, or at least oversimplified view. The real recipe is to be highly qualified on whatever you do and ready to catch an opportunit

I think this is a wrong question, or at least oversimplified view. The real recipe is to be highly qualified on whatever you do and ready to catch an opportunity and ride on an emerging trend. Today the buzzwords are about security, a few years ago it was about services for selfies of teens. On a larger scale there is already biotech bubble and emerging AI bubble, etc. Basically, the strategy should be to follow the money (the demand) and to love what you do (be above average). This, it seems, the most probable way to get noticed, to get funding (for abilities) and to succeed. The markets are stochastic. For example, if you ask yourself, how come that such piles of Java crap as Hadoop came to be so popular, the answer would be that the biotech industry has almost unlimited hot money that time and huge demand for big data processing tools, so even such poorly designed and implemented by amateurs crap would be a good-enough tool. Suppose, I would like to make a similar tool, order of magnitude less wasteful, based on ideas from Plan9, Erlang, based on ZFS, etc, in other works, do it the right way, would I get any funding? No, because there is no real demand for quality solution when a crappy one is OK. There are exceptions, of course, how, for example, nginx became a well-crafted improvement over apache, but this is indeed an exception. So, go to the valley and keep looking. There, it seems, no other way. The principle is that there must be a strong demand backed by big money (Wall Street investors), so even a half-backed result could be easily sold and re-used to return investments and even make some profit.