A very good read in today’s GigaOM provided an impassioned defense of cloud computing. In The delusions that companies have about the cloud author David Girouard, former President of Enterprise at Google, makes the bold statement:
…whether it’s service reliability, data protection, or regulatory issues, there remains to this day an insane resistance to cloud computing that is quickly becoming the “Darwinian litmus test” for companies in every industry.
His “three pervasive dimensions” of this insanity involve fear of outages, the need for customer service during an outage, and a belief that cloud is only good for non-critical apps with non-sensitive data.
Calling it a Darwinian litmus test seems strong, but Girouard comes from a company and a role that gives him remarkable insight into the value and durability of the cloud. In his view, trusting an organization’s most important and sensitive bits to a relative stranger is the only intelligent way to do business. The essence of his argument is that the organizations selling cloud computing have better engineers, higher expectations and more reliability than an enterprise’s own people. In a nutshell, highly specialized beats in-house.
This isn’t a new argument. Companies in the early days of the industrial revolution generated their own electricity before realizing that even occasional outages weren’t worse than the higher cost and similar outages of their own generation units.
Interestingly, Girouard also argues against demanding dedicated customer service and account representatives as it only drives up cost and has no bearing on the quality of service. Going back to the power generation example, no company today expects an ‘inside rep’ to take their call during a break in service.
Natural evolution to cloud
Girouard, with his use of ‘Darwinian litmus test’ makes a compelling point that cloud as the single source for computing power has an inevitability that can’t be ignored. I tend to agree. For startups and companies expanding their footprint, cloud should be the go-to choice.
But for companies with significant investment in infrastructure, it doesn’t make sense to incur the significant cost of a rapid switch to cloud. We’re about to go through a period where companies will be looking to migrate gradually as old systems are retired or where proactively moving computing to cloud can be done without high cost or risk. During that transition, ‘bridge’ services that manage a federated cloud that’s both on and off premise will be a valuable investment.
Whether it will truly be a litmus test remains to be seen, but it will certainly be a standard way to run even the large enterprise in the coming years.