13 years, 6 months ago

All-inclusive Enterprise Architecture

Link: http://www.strategicstructures.com/?p=4

That’s not the most appropriate title. My intention is not to dub another type of Enterprise Architecture (least of all one that sounds like a holiday package). That would defy the points I’m trying to make. Yet I couldn’t help it. The concept of Enterprise Architecture (EA) as I understand it – and luckily I’m not alone – is all-inclusive by definition. It doesn’t mean EA should include everything,  just those things that are important for the enterprise and can be supported by EA management. Some of them are currently missing.

The reason we need EA, according to TOGAF, is to support “delivery of business strategy”. But after that line all listed benefits are solely around IT. The rest of TOGAF and most of the other frameworks are downright IT-centric as well, though admittedly business conscious. Of course they should be business conscious. Just remember what was the driver that brought EA into existence: the gap between business and IT. But it seems the current EA practice is too short a bridge which is now falling down that same gap instead of bridging it. Why? Because it’s not enough “IT” to deliver tangible value and it’s not enough “Business” to make business people really interested. No wonder BPM for example is doing much better on its own (How we came to have EA and BPM as separate disciplines will be the topic of another post). Not that the “B” in BPM is what it should be. Yet, the interest in BPM is growing while in EA is declining, or – to use the Gartner terminology – EA enters trough of disillusionment.

If on the other hand EA was really business driven, not just business conscious, then we should have seen it already in the MBA curricula together with Finance, Marketing, Management and probably right after


There seems to be a broad consensus around EA being strategy enabler. Gartner defines EA as “the process of translating business vision and strategy into effective organizational change”. But then strategy is hardly anywhere in the frameworks and unfortunately not much in the EA practice. TOGAF does not provide any way of seeing how EA supports delivery of business strategy. Actually, motivation concepts were entirely missing until recently. What appeared in the version 9 was a motivation extension in the content framework humbly featuring driver, goal and objective. It’s difficult to understand why motivation should be an extension and not core concept. Shouldn’t it all be derived from there? Other frameworks seem to be doing a bit better in that respect. For DoDAF the concepts of capabilities, measures, and goals (desired effects) play central role in the meta-model and they are well integrated with the enablers – services, activities, people and systems. FEAF in its Performance Reference Model has clarity about value creation from inputs (people, technology, other fixed assets) through outputs (processes and activities) to outcomes (mission, business and customer results).

There were more efforts in this direction in the recent years. And some good results. Notably, one of them is the Enterprise Business Motivation Model of Nick Malik which extended the BMM of OMG, developed by Business Rules Group.

Some good news is coming also from the notations front. A very promising extension of ArchiMate was proposed to fill the motivation gap in the notation and achieve better traceability of requirements. Having motivation in, it’s mostly used to achieve traceability, to have the “why”-chain from the system we implement to what the ultimate business goal that it’s meant to support. Good, but is it good enough? And then, when we need to find that strategic goal, we usually search in the strategic plan. But are all strategies planned? How about “emergent” strategies as Henry Mintzberg called them back in the early 90s. Don’t we see more of those recently? Strategies that emerge from the user experience. Here, user experience. How about that in EA? Or brands? Or probably their place is not in EA. In 2009 Google market value was estimated around $108 billion and the value of its brand over $100 billion. The brand value of most companies in that same list is over 90% of their market value. Do we want to have EA which excludes 90% of the company?

EA needs strategy to direct its change and strategy needs EA to make things happen. And not only for that. Strategy needs architecture for its own design to improve its communication and execution control. After all it was the advent of strategy maps that turned Balanced Scorecard from an original idea to worldwide adoption.


Another general consensus probably first articulated by by Peter Drucker is

Structure must follow strategy

Which structures? Scope wise, here’s where contemporary EA is quite often organisation-centric. EA is about processes, systems and technology. About processes within organisational boundaries. That’s why the real architects of the enterprise, the CEOs, are sceptical about EA value proposition. Because for them

the customer has the process. We have to decide how we want to appear in that process – what we want to contribute to the customer experience – and make sure that’s what our day-to day performance actually achieves

From recrEAtion by Chris Potts

Content wise, we don’t see EA focusing much on structures for non-IT related initiatives. Like for example cost analysis. It’s not new that the actual product and service cost is hidden in the process and cost centre structures. That’s why when bigger and bigger part of the product costs went into the overheads, activity-based costing (ABC) came into the game . EA methods and tools could be an indispensable instrument not only for ABC analysis but much beyond that. And certainly much beyond technology cost optimisation. But OK, let’ have a look now at


I’ve seen many enterprise architecture descriptions. In all of them it seems that the enterprise has only one type of technology – IT. Yes, that’s of primary importance in the service industry. But that’s not the only type. And then, even in cases where investments of non-IT technologies are times bigger than those in IT, those other technologies are entirely missing in the EA blueprints.  Why should you believe in enterprise-level description that is so different from what you read in the balance sheet?

Enterprise Architecture is indeed currently in the “trough of disillusionment”. Yet, it brings a lot of value even as it is. But it can do much better. And it should if it is to get out of the trough. One idea is not to leave out things such as strategy, brand, user experience, knowledge, trust, non-IT technologies. They need EA and EA needs them.