Avoid Cloud Sprawl

Poorly governed cloud procurements and deployments are akin to poorly planned suburban growth: both result in a low-density sprawl that is consequently expensive to service and is associated with inefficient and ultimately unsustainable resource consumption. Cloud sprawl typically occurs as a combination of IT taking a "build-it-and-they'll-come" approach, and business executives making fragmented cloud-based outsourcing decisions without input from IT. Starting discrete cloud buying decisions with a determination of fit based on criteria establish with input from IT is important, but this case-by-case approach not only generally results in much slower adoption of cloud across the enterprise, but may still result in cloud sprawl: micro-level decisions which in aggregate are sub-optimal for the business and which are outside of a cohesive business transformation plan. This presentation will explain how to avoid cloud sprawl by applying portfolio management and the precepts of the Decision Tree (a tool for assessing cloud suitability for discrete cloud buying decisions described in Chapter 3 of the August 2011 book "Cloud Computing for Business - The Open Group Guide") at an enterprise level. In this application, the enterprise's architecture is viewed as a portfolio, and opportunities for cloud adoption are assessed in the context of their relative impact on the performance of the portfolio. Business strategy and business transformation plans can then be modified to deliver the best portfolio performance, and executive compensation can be adjusted to ensure support for well-managed cloud adoption.

- by Penelope Gordon

Author`s Bio:
Penelope specializes in adoption strategies for cloud and other emerging technologies, and portfolio management of early stage innovation. While with IBM, she led innovation, strategy, and product development efforts for all of IBM's product and service divisions; and helped to design, implement, and manage one of the world's first public clouds.