Wednesday, June 22, 2005

Federal Enterprise Architecture

Federal Enterprise Architecture

Over at Business Process Trends, Paul Harmon, Executive Editor/Analyst, and Celia Wolf, CEO/Publisher, do a sensational job of providing news and information related to all aspects of business process change, especially trends, directions and best practices. Their excellent, and free, BPTrends Monthly Newsletter recently discussed some of the innovative work being done by US government agencies as they seek to comply with the Clinger-Cohen act.

The Clinger-Cohen Act of 1996, also known as the Information Technology Management Reform Act, intended among its many purposes to "reform acquisition laws and information technology management of the Federal Government." The Clinger-Cohen Act was implemented to ensure that agencies improve the initial capital planning process for large acquisitions to develop realistic cost, schedule, and performance goals that are tied directly to agency strategic mission goals within available budget resources.

The Clinger-Cohen Act mandates that US government agencies create an Enterprise Architecture "as a means of integrating business processes and agency goals with IT." A number of government agencies are involved in driving the EA initiative, but the leaders are the Government Accountability Office (GAO), the Office of Management and Budget (OMB), the Federal Enterprise Architecture Program Management Office (FEA-PMO) -- which was established by the OMB in February of 2002 -- and the CIO Council, which is made up of CIOs for US government departments.

Now, don't get me wrong. I NEVER LOOK TO THE U.S. FEDERAL GOVERNMENT (or any government agency, for that matter) as a beacon for innovation or brilliant thinking (especially this current Administration and Congress). Indeed, the very fact that leadership for these EA initiatives is emanating from government tends to make me highly suspect that anything of value will ever materialize from these efforts. Nevertheless, I'll try my hardest to give the federal government the benefit of the doubt and overlook my inbred deep-seated prejudices.

To promote the Federal Enterprise Architecture, the GAO has defined five stages of maturity, as follows:
  • Stage 1. Create EA Awareness
  • Stage 2. Build an EA Management Foundation
  • Stage 3. Develop Architecture Products
  • Stage 4. Complete EA Products
  • Stage 5. Leverage EA to Manage Change
Looking at the Federal Enterprise Architecture diagram presented above, you'll see the first component to be developed is the Performance Reference Model. During fiscal 2005 budget reviews the OMB used the Performance Reference Model (PRM) to identify programs that were delivering value and to spotlight others that could not "demonstrate their value consistent with the principles of performance management and measurement."

The FEA Performance Reference Model, pictured below, combines elements of Balanced Scorecards, Baldrige Quality Criteria, Value Chain Theory, plus a number of other additional US government measurement tools designed to provide an overview of the value creation process in the Federal government.

Performance Reference Model

As aptly explained by Celia Wolf in the June 2005 edition of the Business Process Trends Advisor newsletter, the PRM consists of six measurement areas -- each divided into secondary measurements. These include:
Mission and Business Results which measure the specific outcomes an agency seeks

Customer Results which measure how well an agency or specific process within an agency is serving its customers (i.e., US citizens)

Processes and Activities which seek to capture the outputs that are the direct result of the process that an IT initiative supports

Technology which measures the performance of IT in support of specific processes

Two other "placeholders" are also included, Human Capital and Other Fixed Assets, to represent a pair of measurement areas not currently defined in the version 1.0 specification of the Performance Reference Model.
Isn't it ironic that the BPTrends opinion -- which states that "the entire US government effort puts too much emphasis on IT and IT components and not enough on processes and human elements" -- is almost exactly the opposite of my initial impression of FEA?

The longer I look at what the government has created with the Federal Enterprise Architecture, the more concerned I get that this whole thing may have been mislabeled. Perhaps a more befitting name might have been The Omnibus Beltway Bandits Act of 2005. For those of you unfamiliar with Washington DC, let me explain that the term Beltway Bandits is a euphemism that refers to the corps of outside consulting firms who prey on government agencies to the tune of billions of dollars per year.

It worries me how the goals of FEA seem so elusive and ephemeral, especially given the current state-of-the-art of architectural practices (or, more precisely, the lack thereof). It concerns me that so few private sector enterprises have attempted anything even close to the scope or scale being pursued by the federal government at this time. Quite frankly, major, significant, successful EA projects in the commercial business world are pretty much far and few between. Government is trying to leapfrog over industry -- staking out a leadership role in the field of enterprise architecture.

I'm pleased OMB has decided, at least for now, to move forward using a staged implementation approach. I just hope they studied the lessons taught by Sigmund Freud who explained how stages can't be skipped.

I can only imagine how much pressure there must be on the FEA management team to deliver measurable results that demonstrate business value quickly. I just urge everyone to not try and move too quickly. Like the final frontiers of space exploration aboard the USS Enterprise, FEA is trying to go where no IT organization has ever gone before. I just pray they proceed at a measured pace.

One Final Concern

I'm not quite sure what to make of the fact that the Technology Reference Model sits at the bottom of the FEA stack with the arrows in the diagram suggesting that it's the last component to be modeled and populated.

It seems to me the federal government already possess a massive technology portfolio. Granted, I seriously doubt the business leaders who are driving the FEA initiative have an inkling as to what the government actually owns or uses. Furthermore, I imagine these folks haven't a clue how to model or even think about this technology portfolio iceberg that's lying in wait, lurking ahead in the path of their rapidly moving Enterprise Architecture Titanic.

Am I correct in assuming that the FEA proposes waiting until the Performance Reference Model, Business Reference Model, Service Component Reference Model, and Data Reference Model are all developed and populated BEFORE anyone figures out what we can discard from our existing technology portfolio? Don't these managers want to start cutting costs and saving money now? Or, does the federal government prefer to pay consultants exorbitant fees to rearrange deck chairs?

In the computer industry there's a classic joke that goes:
Old Hardware Goes into Museums

Old Software Goes into Production Every Night
The federal government does not have the luxury of beginning with a clean slate. How about rearranging priorities? May I recommend starting with a three step plan: Consolidate. Standardize. Prune. Then go back and iterate repeatedly on all five layers of the Federal Enterprise Architecture.


Post a Comment

<< Home