Friday, September 28, 2012

EA Questions

#entarch #bizarch #UnicomEA Some interesting questions arising at the EA Forum yesterday. I'll post the questions first, and then try and assemble a summary of the answers. (Other participants at the Forum are welcome to chip in.)

Are there any commonly recognized EA priorities? For example, anybody new to EA may not embrace the extent of use of the models we've just heard about, so what do they need to get rightly first, second, etc. as a basis for more sophisticated application of EA?

The panel devoted a lot of time to explaining WHO should lead and perform EA (e.g. not IT, not external consultants) and a lot of time telling people where NOT to start. There was some agreement that you don't start with an EA framework, that frameworks are often designed to solve a particular class of problem, and that you really need to start with a genuine business problem (rather than an IT problem).

I feel like "Enterprise Architecture" has an identity crisis at the moment. I'm hearing "Process", "Decisions", "Systems", but enterprises are far more than that. Where does "geography", "people", "infrastructure" and "customer" feature in an EA model?

There was considerable discussion of Customer-Centric EA, as well as Patient-Centric EA (for healthcare). A number of participants emphasized the difference between Inside-Out EA and Outside-In EA, and pointed out that internally joined-up systems (the focus of some EA frameworks) did not necessarily equate to a joined-up experience for the customer.

How do you validate the quality of an Enterprise Architecture model (in ArchiMate or other)?

There are technical measures of quality, such as consistency and coherence, where tool support may help, but the real measure of quality is in terms of fit-for-purpose. Does this model successfully communicate a given area to the relevant stakeholders, and does it satisfy their concerns?

When mapping motivation against the enterprise, what is the EA role in mapping real stakeholder emotions and impact? 

There may be useful models of motivation and their political implications, but it may be wise to keep these models to oneself.

No comments: