Sounds like a step-by-step legacy code refactoring from an engineer :). I'm sure in the real-world, executing any of the advises will have unknown repercussion

Sounds like a step-by-step legacy code refactoring from an engineer :). I'm sure in the real-world, executing any of the advises will have unknown repercussion. 1. It's hard to fire 10k people _correctly_ and expect things to still work without crushing morale (you may fired potential heroes as well), or perhaps even being poked by government over job losses? 2. I'm not sure a 23 years old hot-shot developer exist. No way Jose. There may be some smart and talented 23 year old, but they're by no mean "hot shot developer". You may get a better result by hiring a 26-32 years old developers. But not fresh grad. Unless your point being to work them like there is no tomorrow, a typical scenario in Silicon Valley startups. Or: today's mess will become tomorrow problem, so you're back to square one with Yahoo! sooner or later => uncontrolled legacy codebases. [Nobody will come out and say that Yahoo has one of the best codebase out there, even Flickr is notoriously bad]. 3. Rounding up the smartest people in the planet is hard to begin with, making sure they all can work together without brushing ego is even harder (especially when everyone wants to leave their mark), finally, expecting rainbows and unicorns to show up is magical I would say. 5. Yeah, whatever, BYOD, use standard toolset, sure. If sys-admin needs their BB, be it. The rest are generics and nothing to complain/argue/discuss. Yahoo! should definitely shed its fat: people and products, no doubt. Next they should think hard on what needs to be done with the successful products they have left. Once stabilization and culture are in place, then you can start doing something more extreme. Rocking the almost tumbling boat seems to be a recipe to drown everybody.