Page tree
Skip to end of metadata
Go to start of metadata

Face2Face at EDUCAUSE 2009 (November 4)

We had 40+ and a full room at EDUCAUSE 2009 in Denver Colorado.

Draft Meeting Notes

Attendees List: will try to get a list of attendees from EDUCAUSE.

The session was presented and chaired by Jim Phelps. Over 40 attendees made it to the session, representing institutions from the USA, Canada, The Netherlands and The United Kingdom among others.


Part 1. Introduction to ITANA (presentation)

Introduction Presentation

Part 2. Interaction

On your campus:
Jim Phelps led discussion in the session looking at current architectural issues being dealt with on the campuses represented at the meeting. A number of themes emerged with supporting anecdotal evidence.

Business Intelligence
Metrics and Indicators, e.g. Financial - reducing costs; metrics used in comparison with peers; metrics used in measures of efficiency.

Business Intelligence discussion - what does BU actually mean. Wikipedia / Gartner comment made as to what it is.
Metrics and indicators - used to gain competitive advantage
Finance <-- reducing costs metrics
Comparison (against peers)
ROI metrics –

Business Processes, Data Management and Architecture

Shifted in to discussion around process. Looking for correlations in the data.
Pointed out the definitions required to fulfil something like 'how many students do you have?' and discussed the difficulty in describing the entity 'student' as a fundamental problem.

Architecture Maturity

Back on problems – how people are engaging with an architectural appraoch vendor driven architecture / customer driven architecture / 'we bought this last week and want it live next week'
Boulder: looking at the best way to look at data architecture. What's the best way to do it for the campus? Budget was cut – question now is top-down the right approach, or can they do some sort of wikipedia approach? So the question is should you go about producing a data architecture.

Washington: one of the reasons they went to EA approach was in reducing print costs - assuming that they could reduce IT related costs from mainframe to switch to IT delivery report. Now basically pushing work out to the consumers of those reports and did not factor in costs/impact. Got push back - asked how they could put that push back to good use in a feedback - how to use in an initiative in something as big as replacing the financial system. How do you quantify the whole impact?

Shift in emphasis from IT to Enterprise Architecture.

Jim Phelps at UW-Madison and the members from University of Washington come from moving up from IT space to EA.

New Mexico reported that in their ERP project, a major positive from taking an architectural approach was achieved in just documenting the current state. Further discussion also pointed out the pitfalls of moving beyond 'just enough' documentation. Continuing to docment current state is also likely to provide diminishing returns if you go too far. The simple rule of thumb here is: "Just Enough. Just In Time."
Therefore there's a balance to be drawn between Rigour and Agility. Reference made to this view being proposed in the current incarnation of the Gartner EA model.

In addition, another comment made was that too producing too many artifacts leads to spending too much time just managing those artifacts. Architect job becomes more a curator role, which has questionable value.

Continuing in the EA as Strategy theme, the group started to discuss the linkages between:

Budget & Architecture & Change

The DoD were used an example of a monolithic approach to EA. The premise of the effort was that it was to be used to control costs by promoting reuse of existing technology, and needing to file for exceptions in the process. However, the monolithic approach has meant that the effort has been 'boiling the ocean' resulting a state where the approach is stifling and not sustainable.

The question was then raised as to what are the right tactics to reach high level goals.
South Carolina looking to lead change via opportunities from EA, as they are currently in the end point of the governance chain as a filter. i.e. by the time they get to hear about initiatives, they are already in-flight.

Washington seconded the point - their programme management / strategic planning are seperate to the development effort.
Where are people on the process? Following the elephant and cleaning up or leading?
Where are the linkages of EA - EA > PMO is one. Other touchpoints include purchasing, budget, etc.

Oren pointed out that the world is changing and moving to disaggregation of systems rather than monolithic systems. This has led to a shift in the definition of 'enterprise'. His suggestion was that there was now, more than ever, a need to undertake "getting out there with the constituents to have the discussion around what 'we' need to have things connected". South Carolina concurred with their own example where they are taking promotion and discussion of SOA to campus.

One observation made was that the discussion primarily, and perhaps unsurprisingly, tended towards discussion of IT Architecture ahead of a broader view of Enterprise Architecture. SURF recognised that their own approach is always to move from the business first. This is also the conclusion of the UK constituents and it is postulated that this is perhaps as much a maturity in IT Architecture modeling as a fundamental approach to EA.

This observation led to discussion of govenrnane models people have in place for EA. An aside to 'Enterprise Architecture As Strategy: Creating a Foundation for Business Execution' by Ross, Weill and Roberston was made in this context, as many in the room have found the book to be a useful reflection point for governance issues.

Jim Phelps, looking at 'EA as Strategy' adoption, highlighted a dffference between failed strategy of commercial entity vs university - where they just come back next year and do it again. One private university countered the point saying that they are publically traded and absolutely have to get te strategy right year-on-year.

Session Closing comments.

  • No labels