It started with an anguished Skype-call from a colleague: “Have you seen this?”, he said. He was pretty upset about it: and after reading the post he’d pointed to, so was I. One of the major consultancies – a self-styled ‘thought-leader’ for the profession – not only promoting methods and tactics that they know are broken and are known to cause significant longer-term damage wherever they’re applied, but presenting them as if they’re something new, a ‘good idea’.

[Although this is about a real incident that’s taken place over the past couple of days, it’s merely one example amongst way too many across just about every industry at present, and it would be unfair to single out just that one organisation. What follows is an enterprise-architecture example, but the same problems apply right across the board: I’ve had painfully-expensive first-hand experience of this from a large accountancy-firm, for example. Several of the players in this case are also people I know and deeply respect: I’ll challenge erroneous thinking, but I won’t attack the person. So I won’t name names here, I won’t provide a link, and although I’ll quote from the original article and comments, I’ll de-identify them somewhat. Fair enough?]

Let’s start with some quotes from the initial article:

Do you want your strategies to succeed? You’ll need the gentleman on the left. He designs your business. … He’s an obscure executive called an Enterprise Architect (EA) and he works for your CIO. … [Y]ou don’t want your strategies following spaghetti roads — you want them moving through your company on logical, straight highways…

Why does he work for the CIO? Because the roads in your company are paved with technology — so the best way to ensure that they are straight is to build and control the tech. Techies invariably screw up the business; business guys screw up the tech. For decades we’ve looked for someone to span both — and that’s what Enterprise Architects do.

Well, perhaps we need to go have a quiet cry in a corner somewhere, because just about every statement in that quote is misleading or wrong – at the least, wrong enough to cause serious damage somewhere. So let’s start from the top, shall we?

First, enterprise-architects do not design the business: they work alongside and with the people who design the business, but that’s not the same at all. To see why we need to be careful about this point, just consider this: the article is saying that some “obscure executive” who “works for your CIO” is the guy who “designs your business” – just how well you think that assertion is going to go down with anyone who works outside of IT?

Second, strategies do follow ‘spaghetti roads’ – that’s the blunt reality, whether we like it or not. By their very nature, strategies are wicked-problems: implementation of a strategy changes the context of that strategy, which means that tactics and plans also need to change dynamically in order to keep track to the aims of the strategy. Strategies also impact on and are implemented by real people – and that’s a whole stream of wicked-problems just in itself. The notion that strategies are predictable tame-problems that can “mov[e] through your company on logical, straight highways” is a product of the kind of Taylorist fantasy of ‘control’ that is guaranteed to fail in real-world practice: is it really a good idea to promote that kind of delusion?

Third, there is a lot more to a business than just its technologies. (To be blunt, the remainder of that paragraph merely repeats the same Taylorist errors: “straight roads”, “build and control the tech”, and so on.) ‘Business’ here seems to mean ‘anything not under the CIO’s mandate’, or, in the classic IT-centric view of EA, ‘anything not-IT that might affect IT’: it ignores the complex relationships and interdependencies between all the other domains across the enterprise, many of which may involve technologies either only peripherally or not at all. Is it a wise move to promote a model that supposedly “designs your business” yet implicitly ignores most of the business of the business?

Fourth, most businesses use many technologies that would not come under the CIO’s remit – for example, see my post ‘The egg-sorting machine‘. For most organisations, the article’s argument actually makes somewhat more sense with the CTO (Chief Technology Officer) than CIO (Chief Information Officer) – but it still doesn’t make sense anyway, whichever domain the EA role is placed in, for all of those reasons above. And yes, I do know that this specific consultancy is trying to rebadge the CIO as having a broader ‘all business technologies’ remit – but it still wouldn’t work, because the EA’s authority would still be constrained to the CIO’s domain. Surely by now people should understand that there’s more to a business than just the infamous ‘business/IT-divide’?

And fifth, the dreaded ‘Naming Problem’: ‘enterprise-architecture’ literally means ‘the architecture of the enterprise’ – so don’t use that term for anything less than a whole-enterprise scope. Enterprise-architectures are only viable when they use whole-system awareness, where everything and nothing is ‘the centre’, all at the same time: we can work within a narrower scope, and in practice we usually do, but that practice must remain aware of the whole at all times. In practice, that systemic-awareness invariably breaks as soon as we place the EA role ‘under’ any specific domain – whether CIO, CTO, CFO or whatever – because the practice becomes centred in and around that specific domain. Given that we know an EA will break if it’s constrained to a single domain, and that doing so will cause damage and/or loss of value for the business, surely we should not promote that kind of approach any more?


But that was just start of it. Unsurprisingly, that consultant’s colleagues dived in to defend him and their firm: and yes, it got worse, as they came up with more and yet more fundamental errors.

For example, we were treated to a classic example of circular-reasoning: their survey of ‘enterprise-architects’ had supposedly showed that 93% of them reported to the CIO, which was purported to prove their assertion that EA belonged under the CIO. But they’d actually surveyed people with the job-title of ‘enterprise-architect’ – without checking what kind of work those people did. If we compare this to the usual ‘title-inflation’ and mislabelling of domain-specialists that’s endemic in recruitment and elsewhere, what we actually find is that almost all of these people are specialist IT-architects – not cross-functional enterprise-architects – and those IT-architecture roles are constrained to a remit that rarely extends anywhere beyond ‘anything not-IT that might affect IT’. In other words, it’s frighteningly like the circular scam that we saw throughout the domestic-violence ‘industry’: here, there’s an arbitrary unverified assumption that it’s only about the business/technology relationship, so support-services are provided exclusively for those concerns, and since only technology-oriented people use or provide those services, therefore that proves that it’s all about technology. In short, not only meaningless, but fundamentally wrong in many different senses – inherently broken, even before it starts – and guaranteed to cause more problems than it solves.

And from there the arguments just went further and further downhill:

What is the most practical approach for the 93% of of our clients that have EA practices whose epicenter is IT? Do we advise them that their EA rolled up to CIO model is hopelessly broken and there is nothing they can do until they fix that (which is a board level decision that many businesses aren’t even remotely ready for)? Heck no…that position is untenable. Do we advise them to have a conversation with their boss (usually the CIO) about the future of ‘architecting the enterprise’ and business technology? Sure, but that doesn’t help them with the problems they have now, it only influences toward a notional future that is still not totally proven.

I repeatedly hear from clients that they come to [our consultancy] because we provide practical answers to immediate pressing problems – so while I think you guys have a point and an idealistic vision about what should be, let’s be good architects and address both the practical, tactical now while keeping the aspirational future square in our sights.

So as with the original post, let’s take this apart in a bit more detail:

— “Let’s be good architects”. Surely the point here is that good architects must “address both the practical, tactical now” with methods and models that are not inherently broken…? Surely it’s not an “idealistic vision”, but a clear business responsibility, to deliver services that actually are fit-for-purpose? Somehow that point seems to have been forgotten there…

 — “Do we advise them that their EA rolled up to CIO model is hopelessly broken?”. Short answer: yes. Don’t beat about the bush on this: it’s broken. Everyone in EA knows it’s broken. Even the strongly IT-oriented Open Group says it’s broken. It really is long past time that we started to be honest about this, and instead tackle the issues in ways that do work.

There are some obvious business corollaries that follow from this. If something is broken, and we know it’s broken, we don’t sell that ‘something’ to others when we know it’s not fit-for-purpose. Apart from anything else, there are serious ethical and legal issues around selling something that we know is broken without advising the client that it’s broken….

Even if the client says they want it, our professional duty-of-care still applies: we must not supply something that we know will cause them damage – especially if that service conceals its own dysfunctionality so as to create an unhealthy addictive dependency on that service. I’m well aware that that’s a very popular business-model for many big consultancies and IT-vendors, but it’s actually the same business-model as a drug-pusher, which is definitely illegal… There are serious business-ethics issues here that are not being acknowledged, let alone addressed: it’s time to stop pretending about that as well.

— “There is nothing they can do until they fix that (which is a board level decision)”. This is simply not true: all those EAs need do is shift perspective, and start doing it properly instead of wrongly. That’s not “a board-level decision”: it’s just a matter of shifting the mindset, from technology-centric to everywhere-centric, and from inside-out to outside-in. It’s not hard to do that shift: but the first requirement is that we have to stop pretending that the old approaches aren’t broken.

— “Heck no…that position is untenable.” ‘Untenable’ from whose perspective? – that’s the point that worries me most. If it’s ‘untenable’ to not sell something that we know doesn’t work, what the heck happened to business-ethics? If it’s ‘untenable’ to not disabuse people of destructive delusions such as IT-centrism and the like, how the heck can we describe ourselves as consultants? If we know that people are driving themselves straight into a ditch with the tactics that they’re using, is it ‘untenable’ to warn them of this? That’s so, so wrong that would take many weeks to explain just why it’s so wrong: but what worries me is that it’s probable the respective people still wouldn’t get it anyway – and yet as Elisabeth Murdoch warned, in a recent speech on the travails of the media-industries, “profit without purpose is a recipe for disaster”. That worries me a lot. Oh well.

Anyway, to sum up:

  • “[the EA] works for your CIO”: that’s IT-centrism – which is definitely broken
  • “the roads in your company are paved with technology”: that’s technology-centrism – which is definitely broken
  • “you want [your strategies] moving through your company on logical, straight highways”: that’s linear-paradigm Taylorism – which is definitely broken
  • “that’s untenable … to advise them that their EA rolled up to CIO model is hopelessly broken”: that’s explicit avoidance of a business-ethics problem – which is also definitely broken

And if it’s broken, don’t pretend that it isn’t – that’s all that we’re asking here. Is that really too much to ask?

2 Comments on “Broken

  1. Its not broken. It never started in the right place.

    IT has rightfully said for years that business does not know what it wants so IT builds as best it can.

    In the businesses where enterprise architecture is effective – it is not advertised. Many such firms don’t even have ‘architecture’ structured formally. Those that do it well and hint at it, keep the IP in house.

    Why would you advertise it – its competitive advantage is huge.

    These may be very large or quite small but it every case it’s about communication – common language, induction, vision to delivery.

    Where is the competitive advantage in common language? Speed, Knowledge and Decision Making based on trust. If your competitors are not moving quickly based on trust and you are – you win.

    • Peter – yeah, good point re ‘competitive advantage’.

      What I meant more was around what Gartner and others have said for some years, that many EA-initiatives have been dropped as ‘failures’. What I’m suggesting is that the reason they fail is when they’re based solely on the linear-paradigm and its notions of ‘control’. Where EA-initiatives do is when the focus is on other themes – particularly around trust, as you say. And yes, in those cases it’s quite likely that commercial organisations would want to keep that kind of success quiet… 🙂

Leave a Reply

Your email address will not be published. Required fields are marked *