I've pondered that very question off and on for several years. I don't know what the "worst" thing would be but some places where I think there are clouds (pun intentionality uncertain) on their horizon - in no particular order: 1) Culture fail: leading to dithering and waste that swamps the ship. Until a recent few years, Google hired and operated like Larry and Sergey's late night bull session dream: hiring "people…
22 points for this? Really? 1) Google hired and operated like Larry and Sergey's late night bull session.. Under Schmidt, et al. in recent years this appears to be starting to be reined in.. You do realize that Schmidt has been there since before they went public, right? ie, he was there for the hiring. They currently have slightly less headcount than at their peak, but they say (in their last earning call no less) t…
2) (Collapse of advertising): Yeah, the on-line ad market ain't growing like it did for a while and in some sectors is contracting. Meanwhile, the "fat middle" of the long tail of the blogosphere is undergoing contraction and professionalization which suggests it is going to evolve into an industry that has far greater leverage in brokering its ads and is not so reliant upon Google. Don't get me started on the coming collapse of the (remaining) major "social networking" sites (which, to a first approximation, have yet to generate a return for most investors other than insiders).
By "decision engine" I mean we'll see ad delivery based less on what a person happens to be reading or searching for, and more on what they explicitly declare to some service that they want to shop for. We've already seen that to some degree in some of the most easily commoditized markets like air travel between A and B, or car insurance, or cars, or electronics parts (http://octopart.com/). That trend is accelerating. Check out a page like: http://octopart.com/parts/category--interface+ics--ethernet/... Aren't the Google ads down the right column of that page just a bit ironic?
3a) (Google's data center potential to fail because of their impulse to monopolize what apps run on them): Ok, well, I have an app that uses (let's say) Oracle Berkeley DB/XML and newLisp atop a GNU/Linux distro of my choice. I'm willing to tweak it for their particular network and storage topology. What's the price?
3b) (overcapitalized): I simply mean that they have to spend more money per quarter than they are competent to spend (or than any other organization with that executive structure could be competent to spend). The same investment dollars would likely perform better if invested with a more diversified set of controls on how they are spent. If you're skeptical of big government, you oughta also be skeptical of big corp.