Re: suggests asking the following questions
For all anyone knows, most of these companies might be hosting their solutions on other company's kit.
To an extent that's barely more or less a problem than having legal ownership of the kit. What counts is whether the company (or nested hosting companies) have a genuinely viable business model.
There's some interesting analysis kicking about the web that reports that half of all AWS phenomenal growth is seriously VC funded startups with no real revenue. And there's quite a few revenue-free tech sector IPO's active in the cloud marketplace - you know of whom I speak. My guess is that the cloud market has already built out to saturation, large segments of demand will evaporate soon as many of the startups implode, or IPOs have to cut their cloth according to their means. What we'll see could well be a re-run of the 2000 telecoms implosion, as cloud providers find that they have little pricing power in a commodity market, and that too much of their costs are essentially fixed as demand declines.
There will certainly be survivors, but its interesting to note that the casualties last time round included obvious companies like Qwest and Global Crossing (at the time with valuations like Amazon today), but it wasn't just the telecoms owner-operators who suffered, but also their supply chain like Nortel and Lucent. When the cloud market implodes, which bit barn & comms equipment suppliers will take a major hit?