Translating Business to Enterprise Architecture: Methodology Activity #4
November 22, 2011 1 Comment
Portfolio Decision Making Methods
The process by which alignment occurs in organizations is never a simple activity. While it would be great to have a mathematical formula that generates enterprise architecture from business strategy, the fact of the matter is that there is no one right answer that fit for all. Having business strategy x does not always translate to enterprise architecture y.
The alignment of Business Strategy and Enterprise Architecture resembles algebra or calculus more than simple addition. The areas that have yielded promise: multi-variant analysis, Real Options (a derivative of options theory); Bayesian models, and quality function deployment. If this appears to be a short course in decision science it is.
In the 80s Marylyn Parker et al, at the IBM Los Angeles Scientific Center developed a methodology, BEAM, which guided MIS directors on the prioritization of I.T. investments. If it was created today it would likely be called Balance Scorecard for I.T. Unfortunately the timing was not right for acceptance of this approach, nor was access to corporate strategy or decision authority granted at the senior MIS levels. But that doesn’t make it less useful today. The heart of BEAM is using multiple weighted criteria to establish a prioritized list of I.T. investments in line with Business priorities.
BEAM worked well in determining priorities; however, in today’s economy long horizon projects such as infrastructure decisions create a problem. Many organizations are unwilling to have projects that have 18 month horizons. Today the name of the game is short cycle and agile. This strategy is useful in limited technical risk, but there is a tradeoff it also restricts the potential upside.
What is needed a means to merge the benefits of large projects with the flexibility or short cycle and agile. This is where options theory shines as a method. The feature behind Real Options (options theory) is the dividing a project or investment into a sequence of investment decisions that are made after over the course of a project. These decisions are whether to invest another segment of the total investment. So at each investment point or option of the project, if partitioned correctly, the decision marker is faced with determining whether the next investment is of value and alignment with current conditions. The previous point is evaluated as to whether it reached its immediate goal, nothing more. Likewise the next investment decision or option is only evaluated in context to what it provides. This decision is made as to whether to invest for what value the option brings or not. A large project becomes a series of continuation decisions over time rather than the big bang investment approach.
Bayesian decision models allow one to yield a decision based upon decision criteria factors weighted as each has different relative importance in a decision. Applying this method –similar to BEAM—enables a more effective decision.
Quality Function Deployment aids in defining the features needed to fulfill needs. This technique can be adapted in the design chain I’ve previously discussed in this blog.