one way of looking at it is, ask yourself how many new products/services has Google launched this year, due to internal development? And then compare that numbe

one way of looking at it is, ask yourself how many new products/services has Google launched this year, due to internal development? And then compare that number to their total employee count of about 22,000. Do you think if say you split off just 50% of those folks, took them and spread them out between hundreds of small little startups each with a fresh and independent entrepreneurial drive and vision, and all legacy process/culture/tradition/rules stripped away, the end result would be way way more new products and services and innovation released into the market? For me, the answer would be yes, I'd expect that. Purely on a people numbers basis, and speaking in general terms. (Of course, it would depend on lots of other factors like the nature of the individuals, etc. But you get the idea.) A software startup only takes 1-4 decent people to get off the ground, and sometimes it doesn't take a heck of a lot more beyond that to scale it up. Now compare that to Google's 22,000. Yes lots of hard-core optimizing of hardware and software going on there, which I respect. But most of them are just polishing corners on Google's legacy products/services, and minding the store, not creating or bringing a fundamentally new and distinct product or service to the market. Also, based on everything I've seen, I think Google hires mostly for highly academic engineering-focused types -- not entrepreneurs and not highly creative types. They mostly get those kinds only through acquisitions (or acqui-hire, as some have called it) lately, not hiring. This is my impression from afar anyway.