Architecture success through standard methodologies

Today was the first day of the Enterprise Architecture Symposium here in Toronto.

Scott Ambler’s keynote presented some survey results showing that Enterprise Architecture programs have been only moderately successful.  Failure is more likely for programs based on a homegrown architecture framework – or no framework at all.

Judith Oja-Gillam and Chris White advocated the pragmatism of picking international standards, such as one architecture framework and one modeling notation, rather than spending time on methodological discussion. Customization to the enterprise is best done by dropping non-essential parts of a methodology.

Ze’ev Ionis boiled it down even further:  for a small or medium enterprise, do a simple Business Function Model, then focus any further architecture & analysis efforts on the “pain points”.

One of the themes at the Symposium was tension between Enterprise Architecture departments and Project Management Offices.  High-quality long-term architecture sometimes gets in the way of delivering projects quickly & cheaply.  Many participants were struggling with internal politics, to get governance structures to balance architecture and project needs.

I will be speaking tomorrow at 2:30pm at the Symposium, about “The Systematic Enterprise: How to synchronize EA, PM and other management methods”.  This will be a holistic perspective on where EA fits in the organization, and will give some ideas for succeeding with any such formal methodologies.