The Enterprise Architecture Network Google Group has soon good disucssions.  Though it sometimes goes beyond what I would strictly call ‘enterprise architecture’; but most EA discussion do.  I would recommend joining it to connect with other architects.  You also have to be a member of the related LinkedIn group.

More for my own reference, here is a contribution I made to a recent discussion on the difficulty in determing an ROI for an enterprise architecture practice:


I think it’s unfortunate when we are asked for an ROI for developing a EA practice.  This is like asking for the ROI of having a project manager.  It’s probably a good question, and I’ve asked it myself, but some things need to be sacred so we can just get on with the job.
I’d never ask for the ROI to set up an EA practice – to me this is sacred – because I’m sold on the idea already, as long as it’s small.  I think we need to sell EA more so less people ask for an ROI. But I also think that means changing slightly how we currently define EA.  It also means we need a response when people ask for an ROI…

This is why I’m so interested in how we define what EA actually is.  I think we should define EA solely as what some on this group might call Enterprise Business Architecture (EBA).  The other deliverables, work products, initiates, etc should be traced to that EBA but we shouldn’t call that the EA.  We also shouldn’t say the EA is complete only when all of the IT components properly align to it.  We should simple do the EBA – showing value streams, key customers, service strategies for the enterprise, etc – and trace to the minimum asset types required to get calculate baseline alignment to the EA.

The reason I say this is because you can’t define a return on investment for developing an EA practice except in terms of improving decision making relating to IT investment.  You are looking to improve alignment to the EBA over time, but you can’t do that from the EA practice itself.  You can only do that by influencing IT investment decisions.  Unfortunately, a more general tool already exists for improving decision making relating to investments.  And that’s the ROI calculation.

The problem is nobody ever asks ‘What is the ROI of determining the ROI?’, nobody ever asks the question ‘What is the quality of the ROI calculation?’, and nobody ever asks ‘How is the ROI process performing?’.  There is a bunch of IT effort being spent which roles up into the ROI calculation and unless you think your IT systems couldn’t be any better at this point in time, it’s not working.

Whether it’s formal or informal, organisations already think they are making decisions to maximize return on investment.  The problem is, in the area of IT investment there is no formal method of determing the inputs which go into the calculation of return on investment.  Also, the inputs must take into account the value streams of the business, IT costs, changes in IT cost structure, and changes in business cost structure.  IT initiatives move costs into IT while creating activities/costs to be performed by users outside IT.  As such, IT can’t determine ROI itself because the initiatives transform more than just the IT organisation.

An EA practice should formalise that process of determining the inputs into ROI calculations. It should also allow the performance of that process to be managed over time.  By the way, within the phase ‘IT initiatives’ I am also including that particular type of project I like which is technology-enabled business transformation.

Other initiatives – business initiatives – simply need to include the costs of IT in their ROI calculation.  IT cannot, by itself, commit to all ‘returns’ which are outcomes of IT initiatives.  This doesn’t mean IT can’t run business transformation programmes.  It simply means that any initative of this type requires communication, traceability, and modeling across multiple disiplines.  And it needs this even to calculate the ROI.

EA, in the strict sense of Enterprise Business Architecture, is the basis for not only IT strategy, but also for IT investment decisions.  EA, in this sense, is the baseline level of knowledge and process required to make ROI calculations for other IT initiatives.  So asking for an ROI for and EA practice which analyses the value streams of an enterprise and traces these to technical and organisational components is like asking for the ROI of developing high quality ROI calculations.

By the way, it’s possible that a particular CIO doesn’t have the responsibility for delivering value from IT investments.  In some instances the IT function doesn’t technically run even IT projects.  Instead these projects are ‘business projects’ relying on IT only to deliver a defined set of IT services.  However, this is a dangerous position for a CIO to be in.  It also just means that the EA practice should be sponsored by somebody else who does have that responsibility.  It doesn’t mean that (a small) EA practice shouldn’t exist.

Matthew De George