Also, EA (whatever framework you use) is an iterative process. You could start anywhere in this circle. As long as you complete a full iteration and preferably a second one, it shouldn't matter where you started out. But you would definitely have had an EA to start with. Most (or rather all) organisations have an EA. Whether they admit it or not, whether it is visible as an EA team or not, is a different matter.
What we are probably referring to, when we mention EA, is perhaps a formal EA team with a defined methodology using an accepted framework and documented fully. IMHO, that is an ideal state, and practically unfeasible and perhaps unrealistic. EA usually never completes. As EA is an iterative process, by the time you reach your goal, it probably might be time to start over again with the next iteration...!!!
Just because you have an EA team using an EA tool, doesn't mean you are actually performing EA.
I would think that having a tool is helpful for completing your EA. But, all your efforts would be towards populating the as-is with lower-level details more useful for support/operations rather than architecture. Not a wasted effort, IMHO. But, I wouldn't want to be spending all my architecture efforts in documenting the complete as-is. I would never hope to complete this task... :-)
Best regards,
Joseph
On May 19, 9:23 am, "Kevin (PragmaticEA.com) Smith"
> first and Tool second" or "Tool First and Architecture Strategy
> second".”
> Using and or starting with a tool is an eminently valid approach to
> start the adoption of EA and EA practises.
> It has been used successfully by many many organisations, for example,
> by Colin Birchenalls presentation at this weeks Gartner EA conference
> in London, where starting with a tool was the absolute best approach
> for Serco and Glasgow City Councils LLP and for the many successes
> they have achieved in growing up to the full EA level by delivering
> massive and real value to the business by modelling using a tool.
> There are many people who have completely failed by starting with a
> tool.
> The reason for this failure is NOT because starting with a tool is
> wrong or bad.
> The reason is (as with many many things that EA seeks to address) is
> process.
> Many people incorrectly start to use the tool (one example of bad
> process but there are many other ways of failing) by modelling the
> heck out of everything and after 6 months (or years for some failed
> attempts) the CEO asks what have we gained and the answer is nothing.
> Don't blame the tool is you use the tool incorrectly. Don't blame the
> screwdriver if you try to hammer nails in with it.
> The successful process for using a tool is actually very simple...
> Stage 1 – Determine the Question
> Identity a high level business question that the management wants an
> answer to, that they either: -
> a) Currently cannot answer, or
> b) Can get an answer but the quality and confidence in that answer is
> too low to be useful.
> Stage 2 – Determine Required Data
> Having understood what the question is, this stage identifies what
> information will be required to answer it. It should be noted that the
> temptation to try to answer the question should be resisted.
> Stage 3 – Populate the Model
> The third stage of the iteration is to find and populate the model
> with the information identified. This sub-process is described in
> detail in the next section
> Stage 4 – Integrate the Model
> This phase ensures that the work that has been done and the
> information loaded into the model is sustainable. For each of the
> Datasources there are two alternatives (which were identified in the
> Analysis Phase).
> a) The Datasource is removed – The processes and people using the
> original Datasource will stop using it and will use the information in
> the model.
> b) The Datasource is preserved – The necessary interfaces are built to
> enable the synchronisation and management of the data going forward.
> Stage 5 – Answer the question
> Having populated the model, it is now possible to use the mode in
> concert with the tools and analyses provided by the modelling tool the
> answer the business question. After this, another iteration is
> possible.
> fyi, I have just started a related discussion the EA Network Linked In
> discussion group (http://www.linkedin.com/groups?home=&gid=36781)
> entitled …
> GARTNER: "A FOOL WITH A TOOL IS STILL A FOOL"
> I was dismayed to hear this same staid, tired and, IMHO, damaging
> catchphrase at this weeks Gartner EA summit in London (May 2010).
>http://www.linkedin.com/groupAnswers?viewQuestionAndAnswers=&gid=3678...
No comments:
Post a Comment