Friday, September 6, 2013

How to get the Wheel Running on EA

Doing Enterprise Architecture is different from managing other type of IT or business project, as it includes from abstract level, logical level to physical level work, perhaps it’s easy to get lost or risk in modeling the universe, so is there any best practices how to partition the EA work and get the wheel running. 
  • Identify principles and a vision for the big picture. DO NOT DO TOO MUCH. Focus on strategic initiatives - that's where the management effort is focused. You would still need a short current state discovery phase to assess, at a high level, the current structure, operations, systems and issues. You may propose a simple evolutionary roadmap - that achieves strategy - the principles and guidelines to be used by other teams to follow the same path.
  • Do not view the EA effort as a single thing. Divide it into the enterprise level and the project or system or solution level. As the lower level, programs are responsible for producing solution or system architecture. It must be harmonious with the standards (including technologies) you identified at the top level. This occurs through a system of governance reviews. Link or attach solution architecture artifacts that have been approved in governance to the top level. 
  • Use an iterative approach that gradually builds a footprint within the organization,  rather than trying to encompass all business areas at once. The partitioning or decomposition of structural and functional elements of the enterprise to accommodate EA will flow naturally. The key outcomes involve identifying which systems are redundant, which databases are redundant, what is missing, what can be reused, and which programs have a business case proving an ROI. 
  • At all times, you have to understand the priorities of the work, communicate again and again, involve people to avoid cultural clashes,  since people either do not comprehend the purpose, the method... and are afraid to ask or just simple forget very quickly. Develop EA opportunistically by taking advantage of what's in the pipeline already. The company won't wait for you to describe the current state in full if they need in the next three months, a high level strategy, roadmap or at least the principles to guide them in what they are doing next. 
  • Break the Silos, and Connect the Dots. The challenge that organization are facing is that most, if not all, of these portfolios are managed in a "silo".  Each silo has its own tools, own methodologies, own frameworks, own people, etc, For example, the IT portfolios that organizations need to manage are in multiple categories, not limited to application, technology, information, change or project, business requirements or demands, the related finances and risk aspects, etc. The challenges to bring them together are obvious. To get the EA wheel running, organizations need to break the silos, integrate the dots between the individual portfolios, and implement automated workflows via a structural framework approach

  • In Pursue of Level 3 -EA as a Decision Influencer. Horizontally, there are three levels of maturity of the relationship between the EA and the business segment leaders. Level 1 exists when the EA is not well known and is not yet trusted. Level 2 exists when the EA is known and trusted but not yet part of the system of decision making. Level 3 exists when the EA is influential in how decisions are made
  • EA is Continuous Improvement Journey. Vertically, there are four levels of architectural maturity of the organization itself. Level 1 is a fairly immature organization architecturally, with limited understanding of the value and role of architecture. Level 2 organizations have some understanding but do not apply architecture in a systematic way. Level 3 organizations have repeatable processes that place architects into key roles of responsibility in planning, alignment, and oversight. Level 4 organizations are continuously improving their architectural practices and place architects into key business decisions.


Post a Comment

Twitter Delicious Facebook Digg Stumbleupon Favorites More