The concept of enterprise architecture is firmly ensconced in most Global 2000 IT organisations (ITOs). Seventy percent of ITOs now have some sort of enterprise architecture (EA) program in place. However, efforts to measure the success and influence of enterprise architecture programs have not been adopted as readily. Fewer than 5% of Global 2000 ITOs with an EA program have a corresponding measurement program in place.
"Enterprise architecture requires measurements to help determine whether the guidance it provides is useful or not," says META Group analyst Philip Allega. "Without measurements, EA is of little use."
EA measurements need to cover at least two areas:
* The impact of enterprise architecture on investment decision-making (eg, IT portfolio management).
* The impact of enterprise architecture inside the project lifecycle, measuring the level of compliance, resistance and the reasons for compliance or resistance.
Enterprise architects currently come up short on both counts. "Investment starts with a project, but it doesn`t end there," says META Group analyst Tim Westbrock. "The decision on when to spend resources on a given project is rarely re-verified or re-examined. The vast majority of organisations do not have an investment planning and prioritisation approval process."
The lack of measurement often leads to what might be termed "intuitive" choices in IT investments based on current fiscal-year considerations. Organisations that pursue such an approach tend to struggle, because IT spending/funding decisions do not take into consideration business performance impact and risk.
"Architects must be able to identify the risk factors associated with making a decision and articulate those risks on a project-by-project and program-by-program basis, all the way up to the macro level," Westbrock says.
Likewise, architects need to maintain visibility within the project lifecycle. "Architects must know how many project decisions were affected as the result of the enterprise architecture program," says Westbrock. "What part of a particular architecture deliverable is having an impact?"
"You want to know who uses architecture and why they use it, but you also want to know why someone does not use it," adds Allega. This information also serves to "close the loop" on the architecture process, creating new input and understanding for the next round of architecture development.
Such measurement programs do not need to be sophisticated, at least initially. "This information can be gathered anecdotally at first. It will be still be helpful in pointing out broad trends," Westbrock says. "The measurements and metrics will evolve as the architecture evolves."
"The question architects need to be asking themselves this year is, `What are you doing to make enterprise architecture more actionable?` Beyond that, there must be a commitment to measuring the success or failure of projects in the context of architectural output," says Allega. "Senior executives want to see measurable results."
USER ACTION: Enterprise architects must track the validity and applicability of architecture deliverables within the project lifecycle and investment decision-making. Measurement efforts must capture the percentage of projects that utilise architecture principles and deliverables, as well as the reasons for and against compliance. Such analysis and feedback is the only way to determine whether architecture guidance is proving useful.
META Group analysts Philip Allega and Tim Westbrock contributed to this article.
Share
