I've actually seen the tipping point between more value in `the cloud' and in our own data-center. The article is leaving out some of the massive overheads in managing your own metal, but even with that factored in, the cloud often comes out to be significantly more expensive.
Finding that tipping point requires that you have a very clear understanding of your company's use-case, which is unlikely to be true when starting out. I often recommend that startups go down the cloud route while still exploring, while keeping a close eye on the costs, and having a clear set of migration criteria and plans.
Comments
I've actually seen the tipping point between more value in `the cloud' and in our own data-center. The article is leaving out some of the massive overheads in managing your own metal, but even with that factored in, the cloud often comes out to be significantly more expensive.
Finding that tipping point requires that you have a very clear understanding of your company's use-case, which is unlikely to be true when starting out. I often recommend that startups go down the cloud route while still exploring, while keeping a close eye on the costs, and having a clear set of migration criteria and plans.