Enterprise Architecture and the 80/20 rule

Gammelgard, Sim0nsson and Lindstrom (2007), argues that enterprise architecture’s scope is wide and the important enterprise architecture frameworks impose extensively detailed models. They argue that too many enterprise architecture initiatives fail because of overly ambitious models. They further state that it is not possible to create enterprise architectures with models in which every detail is recorded. This raises the issue of the reasonable amount of models that has to be created for enterprise architecture descriptions.

According to Koch (2002), the 80/20 rule was discovered in 1897 by Vilfredo Pareto when he noticed regular patterns in the distributions of wealth, irrespective of country or time period. In the distribution, it was always skewed radically to a small minority of top earners with the majority of the wealth (Koch, 2002). Pareto believed that the discovery would have tremendous potential in other fields, a belief that was proven accurate by Joseph Moses Juran in the 1950’s, who proposed that a great many quality faults could be eliminated quickly by focusing on a few vital causes (Koch, 2002). In the 1960’s this principle became known as the 80/20 rule (Koch, 2002).

The effect of the 80/20 rule is that 80 percent of results will flow from 20 percent of causes, and that 80 percent of reward comes from 20 percent of effort (Koch, 2002).

This rule has massive importance for enterprise architecture projects where time and effort is of the utmost importance. In order to rapidly create enterprise architecture descriptions, that 20 percent of architecture that is going to deliver 80 percent of benefit needs to be the focus. Buchanan and Soley (2002), argues that enormous value can be obtained by following this principle in the analysis of business processes and state that the same effect would follow for enterprise architecture development.

Schulman (2003) argues that ‘good enough’ architecture should rely on the principles of:

• Flexibility – separate different architecture domains away from each other in such a way that changes can be isolated and understood in terms of its effect on other architecture domains.
• Most-important pieces – Use the 80/20 rule and build only the minimum of enterprise architecture that will deliver the maximum benefit. Over time the enterprise architecture can be expanded.
• Rapid iteration capability – The 80/20 rule implies enterprise architecture that will change over time as it is added to and amended. Within the boundaries of the architecture principles it should be easy to revise and add onto the enterprise architecture.

According to Schulman (2003):

“‘Good enough’ architecture represents a more-pragmatic view as an approach to an overall architecture concept. The focus is on agility and changeability, with a rapid response to business and technology architecture. By considering the combination of time frame, window and level of effort, a good enough architecture can be created.”

References:

Buchanan R.D., Soley R.M (2002), Aligning Enterprise Architecture and IT investments with Corporate Goals, Available from http://www.bptrends.com/publicationfiles/META%20OMG%20WP%201-15-03.pdf, (Accessed 20 September 2009)

Gammelgård M., Simonsson M., Lindström Å. (2007), An IT management assessment framework: evaluating enterprise architecture scenarios, Information Systems and e-Business Management, 5(4):415-435

Koch R. (2002), The 80/20 Revolution, Nicholas Brealey Publishing: London

Schulman J. (2003), Defining ‘Good Enough’ Architecture, Available from http://www.bus.umich.edu/KresgePublic/Journals/Gartner/research/115900/115962/115962.pdf, (Accessed 20 September 2009)