The Real Problem With Private Clouds

Imagine for a moment you are the general manager of a public cloud. Under your guidance, the company has invested heavily in the hardware, software and the operations necessary to support cloud services. Your market share is increasing, your customers are satisfied, and the broader business is seeing sufficient value that they increase the funding of your division by 35%. This is followed by another increase the next year of 13% - less than you expected, but still an increase nonetheless. But then in your planning for year three, in spite of growth in your customer base, the decision is made to cut your budget by 26%. You have no choice but to make deep cuts, reducing staff and trimming across all budgetary line items. This rollercoaster funding ride continues year after year: +6%, +50%, +6%, -13%, -6%, and so on. By now you are likely asking yourself: who the $*#@ would run a public cloud like that? Well, this example isn’t a public cloud, but the funding of corporate IT within a publicly traded NYSE company over the last eight years.

I’ve written previously regarding all the reasons I do not believe in private clouds: inability of IT to manage to commercial efficiencies, to retain the talent required to manage large scale cloud infrastructure, to compete in industries increasingly converting to consumption-based services, or to keep pace with the technical innovation of commercial cloud vendors. The topic I have not directly tackled is the one most current and former CIOs know all too well: the highly variable funding of corporate IT.

Over the last several years, corporate IT is on a spending spree for new hardware, software and professional services to support their private clouds. The cloud has given IT a second chance within many businesses after years of strategic erosion. CEOs have seen the promise of the cloud and bet that IT can transform their long term competitiveness. This bet is not unlike previous ones, including the web & internet, client server, personal computing, minicomputers, and yes, cloud computing 1.0 (a.k.a. time sharing).

But long after the cloud hype is over, the consultants are gone, and top talent has moved on to greener pastures, these private clouds will need to remain competitive with commercially available solutions from companies whose business models are based on the premise investment is proportional to revenue. Corporate IT will need to extend, enhance and maintain their clouds decades after initial deployment, with funding that can increase or decrease based on the needs of the business and the changing personalities of the senior executive staff.

In the early 1900s before the advent of centralized power, factories deployed and managed their own power plants. As power increasingly became available and accepted as a reliable public utility, the need for proprietary power plants dissipated. Enterprise data centers will some day been seen as the 21st century equivalent of factory power plants, fueled in part by the inability of corporations to justify the funding required to compete against the commercial alternatives.