Distilled from many Enterprise Architecture Consulting engagements, finally published for the EA Practitioners out there to guide their modeling efforts:
- If in doubt, leave it out: if you are uncertain that you need data – leave it out, it simplifies your initiative in the long run;
- Model to the highest level of abstraction, that still gives you value;
- Model to the lowest level of detail, that is still sustainable;
- Figure out over time what the above two levels are for each area and adapt as your maturity and needs change
- Capture stable data, not volatile – your are supporting strategic decisions, not operational needs
- Remember that there is no black or white, right or wrong, but purely better for your organization’s needs;
- If you cannot put it in a visualization or filter of a spreadsheet, the value is questionable;
- Aim first for the overall roadmap and the strategic plan, don’t get bogged down in the details;
- Don’t forget to define standards and principles, because that is what you can easily track (unfortunately the roadmaps are harder to track/audit);
- Start small and lean, evolve over time and don’t boil the ocean and try to build an IT wide data model covering all aspects possible providing all the data needs you could imagine. (Simplify – if in doubt, leave it out)
It is just another list that sparks thoughts, similar to my earlier post:
KarstenVoges - Enterprise Architecture, Startup, Lean 