A subject that popped up a few times at work was Enterprise Architecture and although we seemed to have some expert opinions (yeah right) about it, somehow, no one really had a clue what it was.
In order to find out what EA actually is in practice, I went along to the Enterprise Architecture Summit 2010 organised by Gartner. The event was held on Monday 17th May 2010 in the London Lancaster Hotel, in – you guessed it – London. The event was well organised and had some good content. The Gartner team were professional and very helpful if not a little by-the-book. I went to this event to find out exactly what EA is because I really did not have a clue. As it turns out I was not alone, some of the EA vendors that where in the vendor showcase were not exactly sure either, each seemed to have their own interpretation. One thing for sure was my initial thoughts on what EA is were pretty incorrect.
Sitting in the various Gartner lead presentations it became very apparent that the focus was much more on “Enterprise Business Architecture” which is another way of saying “strategic planning for IT” and much more appropriately describes EA – in my mind at least. It covers the areas of business process and data modelling but at a non-detailed (or non technical) level. There is a lot of cross-over between what EA seems to be and ITIL V3 around service portfolio and service strategy, at least as far as I could tell from my experience working at Hornbill. The main objective behind EA appears to be to allow large enterprise businesses to model the “current” and “future” state of their business systems in order to establish the gaps which are then used to define the strategic projects that need to be initiated in order to get from the current to the future state – all in support of the overall business strategy. In SMB type organisations this normally happens in the head of the visionary driving force, often the CIO or CTO, but when the business gets to a size where a single person simply can not do this for the scale of the business EA steps in.
The EA tools that vendors seem to be offering are essentially repository based planning tools, designed to track versions of artefacts, create or store process and data flow diagrams which are all used to build models of how the business should/could work from a systems and operational process point of view. I was very pleased to see that open source was on the agenda in part, and Wiki’s were talked about as a tool that is often used with EA teams. I am a fan of Wiki technologies for documentation projects, and EA really is on the face of it a big, cyclic, configuration managed documentation project. CMDB’s could be used for this purpose too but in practice a great deal more document management capability would need to be available to make most CMDB solutions a viable EA planning tool.
Despite what generally seems to be thought, EA does not really relate to technology in any direct way at all but it seems to be given to technologists within large organisations, many (but not all) of the people I spoke to their had a technology background but were involved in EA initiatives. As one of the vendors put it beautifully “People think EA is directly related to technology because of the word Architecture, bit it’s actually all to do with strategic business planning that involves documenting and controlling requirements for operational systems, that often (but may not always) get deployed as technological solutions.
EA is all about business planning and creating/maintaining documentation to define what “is now” and what is “to be”. If you consider what a product manager does for a software product, this is really what Enterprise Architects do for internal business systems. EA’s define and plan roadmaps, document them but most importantly communicate over and over again what “is now” and what is “to be” in the future.
There was an abstract presentation that was using biology and the development of biological organisms to relate to systems evolution. It was quite long-winded and I think many people in the room lost it half way through, I was standing because the presentation was full and it was hard to stay there for the hour. The basic message was, rather than designing a whole system from the top down, why not create an environment where the overall business system evolves by localised innovation. Yeah I thought, SOA to the rescue! A reasonable message but surely there could have been a more efficient way of saying it.
A senior analyst from Gartner in a keynote presentation stated that their recommendation for any organisation undertaking a new EA initiative is to not buy any EA specific tools but instead use Office tools (PowerPoint, Visio and Excel or the like) for their initial planning work – it’s all about requirements gathering, definition, management and communications back to stakeholders – for the first 18 months at least. The vendors at the event loved that statement as you can imagine 🙂 Another good point they made is “the role of an EA is really to facilitate rather than design”, the design is an operational function, not one that the EA office should be involved with.
Having spent the time to look into this I came out the other side still not being 100% sure that I fully get EA, but I did figure out it had nothing more than the potential to be influential in relation to the execution of IT projects. For me though, EA is off the agenda for now.
If you know better in relation to EA, please do tell…
This content is published under the Attribution-Noncommercial-Share Alike 3.0 Unported license.
Tibs are pocket-change for the internet™.