I was in the organization (I won't say where) of GTS that came up with the "on demand computing" and "utility computing" concepts. It's rear view thinking at this point, but on a Gartner quadrant chart IBM would have ranked high in vision but mediocre to poor in execution. Believe it or not, IBM had a vision of utility computing that fits in really well with how the reality of cloud computing turned out to be. Had IBM played its cards right, IBM would have been miles ahead of everybody else in the market. However, as everyone now knows IBM didn't play its cards right.
It's only my opinion, but from my perspective IBM differed from the other players (Google, Microsoft and Amazon among others) in that IBM approached the whole utility computing thing as a big integration and marketing exercise rather than something to be built from the ground up. The big challenge facing management at the time was to monetize what IBM had...it's data centers, its hardware, its software catalog, and its legions of outsourcing staff around the world. The pressure was on to "use what you sell".
The result? Amazon was building AWS to automate the management and operation of virtual machines. IBM on the other hand was building websites to take customer orders and pass them on to teams of administrators around the world (lots of India, Brazil and Eastern Europe as I remember, along with various US delivery centers) for manual provisioning and deployment. Throw in lots of DB2, Websphere, and Tivoli products into the mix, because IBM has all that stuff and we should use them.
There were a lot of troubleshooting conference calls...lots of highly credentialed people from every major IBM division, including Research and Tivoli...lots of "lessons learned" and "root cause analysis" chartware passed around between architects, project managers and first and second lines. It would be interesting to see how much of that stuff still lives on today. I can't imagine that much of it survived.
Originally posted by @4twu+1nHiqRRw.