Aggregated enterprise architecture wisdom
By Henry Franken, BIZZdesign Models have played an important role in business for a long time. Process models, information- and data models, application landscapes, strategic models, operational models – you name it, organizations have tried it. …
The next 5 months portend a spectacle of US Congressional battles to be waged ahead of the pending, mandatory “sequester” – automatic, mandatory federal government spending reductions of about $1 Trillion over 9 years, in non-exempt, discretionary appropriations, set to take effect 1/2/2013. Forward-thinking planners in government IT organizations, in large Programs that depend upon IT, and among the Systems Integration (SI) community are likely to dust off Enterprise Architecture skills for analyzing budget cut implications across their IT investment portfolios, and possible cost savings opportunities to offset them. Leveraging a methodical, EA-guided approach to both assess impacts and adjust spending priorities, while illuminating new areas of savings, is a sure route to mitigating serious risks and delays to delivery of critical citizen services.
Whether you’re dusting off the existing EA artifacts, or need to take a very rapid, optimized route to constructing initial enterprise IT models, the driving principle at this time will be rapid, absolute reduction in complexity with a clear line-of-sight to cost savings. “Complexity” here simply refers to an inefficient or needlessly detailed volume of time and resources applied to deliver IT solutions – time spent re-engineering processes, building redundant interfaces and monitors, installing hardware & software in a piecemeal fashion.
Driving complexity, and therefore introducing cost savings, out of engineered systems is the central tenet of “Value Engineering” – “the optimization of a system’s outputs by crafting a mix of performance (function) and costs”. Essentially, deliver the same capabilities with better value by driving down the cost to build and/or operate. Section 52.248-1 of the FAR (Federal Acquisition Regulations) describes the “Value Engineering Clause” that is inserted into many large Federal IT contracts – enabling the contractor to propose changes to the system being developed (i.e. a “Value Engineering Change Proposal”, or VECP). If the proposal is accepted, the actual or collateral savings derived by the government (through cost modification to the contract) can be shared with the contractor. It’s a win-win opportunity for the government, system beneficiaries and the contractor community to discover and propose engineering changes that will lower costs, yet still deliver the same or better results.
Many VECPs are submitted for technology assets – i.e. contained systems that may work better when newer, less-costly components are substituted…like missile systems or electronic devices. This may be because the contractor typically is the sole source of knowledge and research concerning how to optimize and build the components, the “value” and function of the asset is very clear (i.e. it’s delivered and explodes on target) and constant innovation is a demand of the environment. VECPs are also submitted for IT systems and programs, though it’s much more difficult to identify and propose the specific cost savings or cost avoidance that might result – since IT systems are frequently dependent upon many external or interfaced elements, vendor products and processes.
An EA-centric review of a program or line-of-business IT investment may quickly yield insight that would lead to specific value engineering opportunities, and therefore reductions in IT costs. For example, a particular set of information may be created, shared and recreated across several systems, using different processes, datastores and technologies. An segmented EA approach driving down from the particular business, process and information domains, may quickly illuminate targets of opportunity for database or interface consolidation – and therefore potential consolidation of supporting technologies (i.e. storage, networking, processors). This may lead to optimized technical operations and business process performance, which can be clearly mapped back to the Enterprise Architecture to validate that governance and mission requirements are (still) met, cross-enterprise risks are mitigated, and IT investment portfolios and procurement activities are properly adjusted or re-aligned.
With this kind of information, a VECP could be constructed that very clearly proposes both program-specific and collateral (i.e. across the rest of the enterprise) savings resulting from introduction of state-of-the-art consolidating technologies (for example, pre-integrated, self-contained and consolidated database engineered systems, perhaps cloud-enabled). At the very least, an EA view can help identify and prioritize targets of opportunity for Value Engineering that may become part of an effective and timely sequestration response – both before and after such an event, and in fact as part of the annual capital planning and investment control (CPIC) processes.
The next 5 months portend a spectacle of US Congressional battles to be waged ahead of the pending, mandatory “sequester” – automatic, mandatory federal government spending reductions of about $1 Trillion over 9 years, in non-exempt, discretionary appropriations, set to take effect 1/2/2013. Forward-thinking planners in government IT organizations, in large Programs that depend upon IT, and among the Systems Integration (SI) community are likely to dust off Enterprise Architecture skills for analyzing budget cut implications across their IT investment portfolios, and possible cost savings opportunities to offset them. Leveraging a methodical, EA-guided approach to both assess impacts and adjust spending priorities, while illuminating new areas of savings, is a sure route to mitigating serious risks and delays to delivery of critical citizen services.
Whether you’re dusting off the existing EA artifacts, or need to take a very rapid, optimized route to constructing initial enterprise IT models, the driving principle at this time will be rapid, absolute reduction in complexity with a clear line-of-sight to cost savings. “Complexity” here simply refers to an inefficient or needlessly detailed volume of time and resources applied to deliver IT solutions – time spent re-engineering processes, building redundant interfaces and monitors, installing hardware & software in a piecemeal fashion.
Driving complexity, and therefore introducing cost savings, out of engineered systems is the central tenet of “Value Engineering” – “the optimization of a system’s outputs by crafting a mix of performance (function) and costs”. Essentially, deliver the same capabilities with better value by driving down the cost to build and/or operate. Section 52.248-1 of the FAR (Federal Acquisition Regulations) describes the “Value Engineering Clause” that is inserted into many large Federal IT contracts – enabling the contractor to propose changes to the system being developed (i.e. a “Value Engineering Change Proposal”, or VECP). If the proposal is accepted, the actual or collateral savings derived by the government (through cost modification to the contract) can be shared with the contractor. It’s a win-win opportunity for the government, system beneficiaries and the contractor community to discover and propose engineering changes that will lower costs, yet still deliver the same or better results.
Many VECPs are submitted for technology assets – i.e. contained systems that may work better when newer, less-costly components are substituted…like missile systems or electronic devices. This may be because the contractor typically is the sole source of knowledge and research concerning how to optimize and build the components, the “value” and function of the asset is very clear (i.e. it’s delivered and explodes on target) and constant innovation is a demand of the environment. VECPs are also submitted for IT systems and programs, though it’s much more difficult to identify and propose the specific cost savings or cost avoidance that might result – since IT systems are frequently dependent upon many external or interfaced elements, vendor products and processes.
An EA-centric review of a program or line-of-business IT investment may quickly yield insight that would lead to specific value engineering opportunities, and therefore reductions in IT costs. For example, a particular set of information may be created, shared and recreated across several systems, using different processes, datastores and technologies. An segmented EA approach driving down from the particular business, process and information domains, may quickly illuminate targets of opportunity for database or interface consolidation – and therefore potential consolidation of supporting technologies (i.e. storage, networking, processors). This may lead to optimized technical operations and business process performance, which can be clearly mapped back to the Enterprise Architecture to validate that governance and mission requirements are (still) met, cross-enterprise risks are mitigated, and IT investment portfolios and procurement activities are properly adjusted or re-aligned.
With this kind of information, a VECP could be constructed that very clearly proposes both program-specific and collateral (i.e. across the rest of the enterprise) savings resulting from introduction of state-of-the-art consolidating technologies (for example, pre-integrated, self-contained and consolidated database engineered systems, perhaps cloud-enabled). At the very least, an EA view can help identify and prioritize targets of opportunity for Value Engineering that may become part of an effective and timely sequestration response – both before and after such an event, and in fact as part of the annual capital planning and investment control (CPIC) processes.
For enterprise-architects, the Olympics provides a great illustration of the difference between ‘the enterprise’ and ‘the organisation’. Although in business it’s quite common to regard ‘enterprise’ and ‘organisation’ as synonyms for each other, for enterprise-architecture its wisest to use the…
Many companies do not do governance well. A primary reason for this is a focus on governance “process” at the expense of policies. And, where policies are established, it is common to observe a surfeit of bad, inconsistent policies that are overlapping and generally ignored. As a result much governance is carried out by opinion; and governance decisions are not easily repeatable.
The Cobit 5 framework provides reference models for process and goals but, other than providing very general guidance, stops short of any detail at all relating to principles and policy. However in fairness Cobit 5 does recommend “a (hierarchical) structure into which all policies should fit and clearly make the link to the underlying principles”.
So what does a policy hierarchy look like? Does each organization need to invent its own unique structure and content? Actually we need more than just a policy hierarchy, we need a model that helps us establish a consistent approach to policy search and description. And whilst every organization will have unique needs, much of the hierarchy and policy content will be reusable. What will usually be highly customized are the contexts and their relationships with policy assertions.
In the diagram:
– policy type – classifies the policy. It can be hierarchic.
– policy subject – identifies the focus of the policy the class of object being governed.
– policy – a strategy or directive defined independently from how it is carried out
– policy assertion – is an atomic policy requirement, expressed as a statement that must be true or false
– policy context – an entity that limits the reach of a Policy.
– policy effect – an intended and/or an actual outcome of a Business Policy. This can be the Principle(s), Goal(s) or Outcome(s), which of course map neatly to Cobit 5.
Let’s look at an example:
| Meta Class | Example |
|---|---|
| Policy Type | Architecture |
| Policy Subject | Application Architecture |
| Policy | Interfacing |
| Policy Assertion | All new Application Interfaces must be loose coupled. |
| Policy Context | Global applicability |
| Policy Effect | Principle: Interoperable; IT Goal: Agility |
Now to put this more broadly into the Cobit 5 context, here’s a fragment of a policy hierarchy, mapped to Policy Subect and Cobit 5 IT Goals.
The policy hierarchy shown above is not rocket science. However it facilitates consistency and communication to all the various stakeholders. You could at a stretch manage policies in a spreadsheet, but in practice it would be advisable to use something like Sharepoint or an equivalent, that allows you to manage the life cycle, status and so on. In a further elaborations of this little series of blog posts I will explore policy relationships with guidance and standards, policy assertion and context development plus the broader policy management model.
Reference:
Using Cobit 5 – Part 1: Principles
Using Cobit 5 – Part 2: Policy Nomenclature
Next Step: Talk to David about how to apply effective, policy based governance.
As the cloud computing adaption and maturity accelerates, a number of case studies of early cloud migration are emerging. Ironically most of such case studies often talk about success of such migration and dynamic business and technology benefits it de…
As the cloud computing adaption and maturity accelerates, a number of case studies of early cloud migration are emerging. Ironically most of such case studies often talk about success of such migration and dynamic business and technology benefits it de…
by: Bill Cason – Troux CTO – July 31st, 2012
Gathering and assessing data can be quite seductive. But it’s the equivalent of modelling the
universe, and it’s a recipe for disaster. In fact if we see any problem today in our EA deployments, it’s that people get so excited they want to gather all their data at once. This is the last of the Top Three common mistakes Enterprise Architects (EAs) make when starting (or re-starting) a program. I call this “Building the Answer Machine” – and it doesn’t work.
In this scenario, you are asking people for more data than you need, without any scope control or business value focus. Obviously you don’t want to collect data, just to see what you can find.
Instead, consider these three recommendations when embarking on a data gathering effort. They will help you successfully identify and gather the data that is most important to your business. #1- Leverage a wider team with a focus on business value to help you identify which data is critical. #2- Automate as much of the data collection process as possible. #3- Market your success internally so that contributors appreciate the fruits of their labor and are more inclined to embrace the data gathering effort moving forward.
Leverage a wider team
You will have lots of stakeholders and constituencies requesting answers. You can quickly lose sight of being able to deliver value in a timeframe that is acceptable to management. Queuing up priorities and managing scope control is hugely important. Otherwise you will have plenty of stakeholders disappointed by your inability to answer their questions.
The EA team is responsible for creating business value with the information that is collected. But don’t waste EA resources gathering that information. With executive sponsorship in place, the enterprise will see it as a priority to identify data stewards. The data stewards will then provide the required data you could not automatically collect. This in turn ensures EA resources are focused on analyzing the quality of the information, and identifying gaps in order to initiate the decision-making process.
Automate data collection process
Based on my team’s experience, as much as 80 percent of the information required to initiate an EA program already exists in the enterprise. In many cases, you can automatically collect important information from other IT and business planning repositories so you don’t have to expend human resources to find what’s already being managed. That said, stay focused on collecting only the information necessary to answer the high priority questions your business wants to address.
Market your success
Don’t forget to market your success internally. You secured support from the organization by promising something good for them, so make sure you go back and tell them you did it. Then the organization as a whole can share in your success.
When you gather all this information and start to see results – in this case the answers to critical business questions – share those results.
Remember, the data acquisition process is accretive. The data you get to answer the first set of questions becomes foundational for answering the second set of questions. You don’t use information, throw it away and stop asking questions. By involving the wider team, you empower and encourage people to embrace the EA processes and use the output to change the business.
We already know that organizational change is one of the hardest challenges any company can embark upon. Ensure you are taking the right steps by aligning a wider team to provide information, automating the data collection process, and marketing your success internally. The EA practice can then deliver true organizational change in a focused and organized manner, on a timeline management expects, and with the support of both your executive sponsor and the whole of the enterprise.
Read other articles in this three-part series: How to avoid common mistakes with your EA Program:
The previous post on ‘Money, price and value in enterprise-architecture‘ was kinda long, so here’s a (somewhat) shorter summary: Background It’s fundamentally important that enterprise-architectures should incorporate the following assertions: there are many other forms of value besides money in…
To my mind, the way that most people seem to handle the relationships between money, price and value in enterprise-architecture and EA models is, frankly, an utter shambles. So let’s get a few things straight here: there are many other…
Continuing on from ‘Framework versus body-of-knowledge‘, the same colleague asked me for some notes on how we could apply quality-systems concepts to enterprise-architecture itself. Background to this is that perhaps a dozen years back, I was working at an engineering…
Erroneously, I always believed that Art did not belong in Enterprise Architecture – that what was required were clear, unambiguous, clear and concise designs – that was until I came across the Dragon1 way of approaching Enterprise Architecture www.drag…