Guest Post: The Value of Strategic IT Planning and Enterprise Architecture

Did you know that 80% of all enterprise architecture (EA) initiatives are not completed because they fail to demonstrate “value added” to current business practices?  I recently attended the Troux Directions Federal Users Conference, which made some excellent points on the theme: “How does EA add value?”

Not only was I impressed with the level of market awareness articulated by the Troux staff, but the government and corporate presentations offered some interesting insights. Here are some take-away notes from the conference that may help with your Strategic IT Planning (SITP) initiatives and show you how a mature EA can add value and perhaps generate some continuing dialogue. 

Presentations:

• The Department of Homeland Security (DHS) presented their EA methodology and how they use Troux for their strategic IT planning efforts. They are moving onto a phase where the enterprise architects are collaborating with their strategic planning office, which is consistent with the notion that EA and SITP are inseparably interrelated. DHS also has plans to move to a Federated Architecture, which they hope will create a movement in the EA and IT space, and allows them to share information between agencies and organizations. To this end, they willingly share their constructs or data structures with other agencies and industries. They presented a great template for how you can gather valuable information. 

• In addition, there were many more briefings on EA value-added by the USMC CIO office, the Department of Health and Human Services (DHHS), EDS, CA, SRA, and ManTech, which added to the following list of observations.

The EA Landscape:

• Developing an enterprise architecture using PowerPoint presentations and Excel spreadsheets are a thing of the past. Sophisticated tools are available and uniquely positioned to bring IT planning to a new level of strategic implementation and forecasting.  If you find yourself burdened by regular use of homespun spreadsheets to manage large volumes of information, with multiple tabs and hyperlinks to documents, other spreadsheets, websites, presentations, etc. to access information, and then have to manually digest and analyze the information into a workable presentation for decisions on the enterprise, you are overdue for adopting more advanced EA and SITP tools for automating these same processes.

• It may take several years for large agencies to adopt an EA approach to strategic planning or create a culture that is EA-minded. Therefore, EA customers need something they can deploy quickly, get value from fast and often. Business intelligence needs to be transformed thorough the EA into intelligent business practice. The reality is that it takes time and resources to generate a useful model so agencies can learn from one another’s best practices, successes, and failures.

• The use of SITP is on the rise as evidenced by more than  200 participants to this User Group Conference, and agencies mentioned more than once a need for federated modeling and data standards so they can share data when appropriate (i.e., crisis response) between agencies.

• EA teams are relatively small (typically five to six principles) regardless of the agency, and included operational, logistics, engineering, modeling, financial and architectural savvy team members.

• Considering rice bowls, inter-agency cultures, and other inherent roadblocks, EA requires an effective champion to produce or obtain the data, mandate the process, or influence and make budgetary decisions regarding its use. Be sure to find a champion first. Someone must see the value added and is willing to foster EA initiatives.

EA Value Added:

• Quick success may ensure the value of EA initially, but the long-term value is in using EA for Strategic IT Planning. You can achieve both simultaneously, which will be important to your champion; your advocate and resource sponsor.

• CIOs, Program Managers, Capability Portfolio Managers, operators, engineers and financial management personnel can customize their own dashboards to have daily insights from the EA, according to their own preference or perspective. Basically, they all run into similar problems like access to information, relevancy, integration, etc.

• “Buy-in” from agencies contributing data or users of the EA is absolutely essential to its success.  Authoritative and reliable sources of information need to be identified early in the build process and then partnered with for the success and credibility of the enterprise.  However, be prepared to cross cultural boundaries between organizations as you identify and evaluate their contributing potential to your EA and SITP initiatives; not everyone likes to share.

• The EA data refresh mechanism must be automated to the greatest extent possible. If not, data is less likely to be kept up- to-date and then it becomes a more burdensome governance issue. If someone is not generating the data, don’t get into the data generating business. EA will never replace the man-in-the-loop since not everything can or will be automated, especially analysis.

• Optimum performance of the EA would include maximum use of automated, cyclical data mining of existing databases (reliable, verifiable, and authoritative). Many relevant databases exist and can be tied to architectures with relatively little effort – push and pull capability is available now. Your real challenge may be finding authoritative data sources that are also reliable – they are not mutually exclusive.

• There are also security and access issues that must be addressed up front. When does the collection of data become classified?  Who has or controls access to the data? Where does the data reside?  There are no security classification guides to my knowledge that spells out when the volume of information in the enterprise compromises the security of the enterprise; unless of course the data in the enterprise dictates the security level.  You need to be cognizant of that turning point and act with prudence and foresight. 

• Generating data and views from existing data is standard business practice. Relating disparate information in the EA environment lends itself well to clearer understanding.  Although not discussed in detail, obtaining useful SITP information from architectural artifacts has produced some success in this area.

• Visualization of the enterprise (programs, processes, data, relationships, assessments) is supremely important and quickly adds value to enterprise partners by keeping everyone on the same page during dynamic evolutions of the architecture.  While there may be an immediate ROI with “out-of-the-box” capabilities in reports, dash boards, etc., one view does not fit the needs of all of the stakeholders.  An application’s ability to customize the front end view to suit the needs of the echelon stakeholder is supremely important to getting the maximum value out of the EA.

• The EA needs to talk the language of the agency, business, unit or level of client it serves.  The objective of EA tools is to deliver purposed information to a variety of stakeholders.  You will need to customize the views for it to be useful and used or else it will be an exercise in futility.

• Metadata standardization, interoperability and artifact reuse not only saves time and resources, but opens venues for federated use of architectural information between agencies using different tools.  For this to effectively work, data latency and business rules need to be identified, thoroughly vetted and understood through metadata tagging.  This is also key to maintaining a living architecture and for accomplishing accurate analysis within your enterprise.

• Time and resources spent on the data’s pedigree up front will help establish the credibility of the architecture as an authoritative source for enterprise partners and participants.  The federation’s success is dependent on accurate, timely and relevant information from the enterprise and an established, verifiable pedigree from whence it came.  If you plan to tap into another enterprise’s data sources, the mechanism for refreshing and/or maintaining its vitality must also be considered.

• EA will continue to grow at its own pace until legislation mandates its use for IT Planning – then it will be grow more exponentially. Just like the cash register wasn’t popular until congress changed the laws to value the receipt as a legitimate contract and for tax return claims.  EA policy guidance and reporting requirements already exist.

• A mature EA provides real-time enterprise information to multiple stakeholders delivered in user specific views; via a central repository. It also provides traceability of capabilities to requirements, and allows investment analysis, decision support, identification of cross component capabilities, and performance analysis.

• Definition of an EA USE CASE: Information the users are interested in knowing. To determine what information is paramount to the EA, questions related to governance, business processes, data (needs, existence, organization, source, etc.), analysis views, technical, programmatics, resourcing and functionality need to be addressed across the enterprise. Users of the EA may be small in number or diversely large dispersed geographically so the EA Framework must be useful and visible to all potential users.

• No one can learn everything about the enterprise and not everyone is an enterprise architect with their hands on the data from day to day, so you need to know what you’re looking for and how to get to it. EA in a metadata rich environment provides one mechanism to locate, assess, display and use authoritative, pedigreed data for more informed decision making to the nth degree. EA and SITP tools are designed to facilitate that end.

Summary

Agencies and industry are most interested in the bottom-line of EA – what value-add can be derived from its efforts. The ROI must show that EA will outweigh the expense of creating and maintaining it. A fully functional EA may take several years to accomplish, so to maintain advocacy, the EA and SITP teams must quickly show the value add.

Ultimately a mature, well-administered EA has the potential to facilitate a vision and affect positive changes in programs, policies, resources, and schedules, providing unilateral continuity, and saving time and resources while building a more capable enterprise.  If done right, the enterprise architecture used for strategic IT planning will help answer questions before they are asked.

About the Author

Vinny DiGirolamo, AFCEA Senior Fellow and author of Naval Command and Control: Policy, Programs, People and Issues (AFCEA International Press) is founder of Capital Investment and Technology Consultants (www.citc2.com) and has facilitated multiple EA and SITP initiatives for the DoD and industry.