I was talking earlier on to one of the people we interviewed for the book, outlining the main findings and getting some feedback. One of the main topics we focused on was the nature of Enterprise Architecture, and what characterises EA work “done right”.

The person I was talking to relayed a conversation they had with a client recently that was interested in becoming “Zachman compliant”. The client was under the impression that if they could do this, it would be a stamp of approval that they were “doing EA well”.

I’ve seen things like this myself from time to time, and it’s worrying. Organisations set up EA teams that are naturally inclined to focus on modelling technology landscapes and so on, and that focus more on the outputs (models) than on how they arrive at those outputs.

The problem here is that the process of EA is at least as important as the outputs. “EA done right” is as much about how you conduct EA work, as it is about what you produce. EA done right involves architects engaging in truly two-way discussions with stakeholders in IT and business—sharing ideas, learning and also communicating what’s been learned so far, so everyone understands more. EA done wrong involves architects sucking ideas from the brains of “users” and not giving anything back. The architects learn more but very quickly you end up with a fundamental disconnection between business and IT and you’re no closer to developing a common language.

And this brings us back to Zachman—because although ZIFA goes further than just promoting the Zachman framework (it’s starting to promote the importance of developing a good process) there are an awful lot of people out there who just focus on the framework. And although the framework is a great way of categorizing models and knowledge, it says nothing about how that knowledge is arrived at.

If you just focus on getting good at filling in the cells in the Zachman framework, that’s no guarantee at all you will be doing EA in a way that helps IT or the business.