This one starts from a question from Arif Mustafa, an IT-strategy and enterprise-architecture consultant with Toyota Canada, over in Toronto:
Why (newer) enterprises like, say Google, Facebook, Amazon etc are able to create culture of collaborative-innovation where companies like IBM fail to do so. Does it have to do with (among other things) organizational structure and managerial style? I am interested in understanding the influence of “Organizational structure” on success and sustainability of the enterprise – if that were truly the case, why can’t enterprises simply restructure them for success.
The quick answer is that changing the structure on its own isn’t likely to make much real difference – if any at all, in the longer run. To get real change, we need to change not so much the structure as the story – the deep-story or deep-myth that underlies the surface-level ‘rationality’ of the overall organisation.
The point here is that the organisational-structure of an organisation tends to reflect its underlying story: if we try to change just the structure without changing the story, the structure will revert back automatically to some (possibly-covert) version of its previous form, often surprisingly quickly. Robert Pirsig described this point well in chapter 8 of his now-classic book Zen and the Art of Motorcycle Maintenance:
The true system, the real system, is our present construction of systematic thought itself, rationality itself, and if a factory is torn down but the rationality which produced it is left standing, then that rationality will simply produce another factory. If a revolution destroys a systematic government, but the systematic patterns of thought that produced that government are left intact, then those patterns will repeat themselves in the succeeding government. There’s so much talk about the system. And so little understanding.
Hence, for example, we might well desire to change the overall structures of the organisation to support the type of market-relationships described in the previous post, ‘Keep the focus on the right market!‘. In visual form, we would aim to place ourselves and our organisation within a shared-enterprise type of model:
And, from there, we would view the organisation itself, and all of its subsidiary business-functions and the like, as sets of services within the overall frame, guided by the Enterprise Canvas mapping of interrelationships and interdependencies between those services:
But if the core paradigm and worldview of the organisation and, especially, its management, is essentially Taylorist and neo-feudal, the organisational will automatically revert to that type of structure. We can decide to re-organise to a matrix-structure, to a podular structure, to a holacracy-structure, to ‘managerless teams’, or whatever: it doesn’t make any difference, really. If the story doesn’t change (such as if, for example, as Paul Sutton put it in a recent Tweet, there’s a deep assumption that “‘Bosses do the thinking & workers wield screwdrivers – You’re convinced deep down that’s the right way to run a business.’ Matsushita”), then what we’ll get, sooner or later – and usually sooner rather than later – will be this:
Which, as we saw in the ‘Focus on the right market’ post, doesn’t work – or work well in most current business-contexts, anyway. But people keep on coming back to it - because the deep-story doesn’t change.
The catch, of course, is that changing a surface-level structure is (relatively) easy: that’s all that a ‘reorganisation’ is, after all, and many companies seem to undergo that kind of pain almost on a continuous basis…). But changing the story is not easy at all…
I replied to Arif along those lines above, and he came back with the following:
I understand, just structural change without cultural change makes any sense – but one could flip that argument that organizational structure is reflection of an organization’s culture. I was wondering if there are optimal structures (based on culture).
I’m always a bit wary about ‘optimal’, because a true ‘optimum’ is always going to be context-dependent, and probably somewhat fleeting at that. Somewhere there’s some eponymous ‘law’ or principle about this, that structure reflects deep-culture – a variant of Conway’s Law that the structure of software reflects the structure of the organisation that created the software, that “organizations which design systems … are constrained to produce designs which are copies of the communication structures of these organizations”. As Gene Hughson noted in a recent post, it’s also elegantly expressed in Ruth Malan’s paraphrase of Conway’s Law: “if the architecture of the system and the architecture of the organization are at odds, the architecture of the organization wins”.
(I seem to remember Dave Snowden once saying much the same in another context, something about “when two companies join, the company with the most bureaucratic culture wins”. It’s one of the reasons why Amazon had to be so careful to maintain a ‘hands-off’ relationship with Zappos after they’d acquired it, because the whole of point of the acquisition was for Amazon to learn from Zappo’s culture – which would have been squashed, maybe within days, if they’d simply subsumed it into the larger organisation.)
The real key to this – to make it work – is a combination of awareness, choice, and dogged persistence to maintain both of those. Pixar’s Ed Catmull gives a great real-world example of this in a recent FastCompany interview, ‘Pixar’s Ed Catmull on why communication should know no boundaries‘:
The managers were “trying to carefully keep this thing from going off the rails,” Catmull says. Reining in the animators and artists for the entire production, in a high-risk project for the company, they followed the proper protocol to communicate within the company. But they ended up feeling under-appreciated; like “second class citizens…sand in the gears.”
Communication, says Catmull, should be separate from the structural hierarchy of an organization–out of order, and among anyone within the company, anytime. That may sounds like chaos to managers, whose job it is to keep the team coloring inside the lines.
“You’ve got to get over it,” he says. “If you want people to be let loose, you can’t over-control them.”
(By the way, there’s a lot more detail on this in Ed Catmull’s recent book Creativity, Inc – strongly recommended for anyone in the business-architecture / whole-enterprise-architecture space.)
All of which is fine – but most organisations are not Pixar. So how do we do this in practice?
Practical implications for enterprise-architecture
There’s quite a lot of mainstream business-literature around this: Kim and Mauborgne’s 2009 HBR article ‘How strategy changes structure‘ is one of the more useful ones, for example. The catch is that way too much much of that literature – in the mainstream anyway – tends to be a bit too full of the delusion that we can change culture by executive diktat: which, of course, we can’t. Or rather, the executive diktat can go out, with force of law, if need be: but in reality, such diktat only acts at the surface level, and rarely reaches down into the deep-story – not merely the espoused values, but the enacted values – which is where real change happens.
To get to the deep-story, we need tools that are specifically designed to elicit it and bring it to the surface. One I would always recommend for this, even for assessment of seemingly-straightforward inter-stakeholder relationships, is Nigel Green and Carl Bate’s VPEC-T – Values, Policies, Events, Content, Trust.
For deeper work, especially at larger scope and scale, I’d recommend Sohail Inayatullah’s Causal Layered Analysis (CLA). It was originally designed for use in futures-work, and some technology-oriented folks might struggle somewhat to get the head properly around the way it works – its background is more in linguistics and the like, as indicated by its subtitle of ‘postructuralism as method’- but it is worth the effort. To get you started, perhaps, I’ve written quite a bit about using CLA in enterprise-architecture, such as in these three posts:
- Causal Layered Analysis, SCCC and Cynefin
- SCAN and Causal Layered Analysis
- Two SCAN notes – 2: Causal Layered Analysis
Yet that’s only the ‘awareness‘ side of the tasks: the other side is about the ‘choices‘ – about how we change the culture, more towards a shared-optimum that works well with the broader shared-enterprise. And we do that by changing the story – not at the surface, but a much deeper level.
Over in the US, some of the people whose work I know and admire on this include Michael Margolis (GetStoried), Michelle James (CreativeEmergence) and Mike Bonifer (GameChangers). The person whose work I know the best, though, would be Shawn Callahan and his colleagues at Australian-based Anecdote: for example, see Shawn’s recent post ’One simple way to bring your stories to life‘.
The key in all cases is to leverage ‘small-stories‘ – much more so than grand Hollywood-style ‘Hero’s Journey’ stories – to guide subtle changes in the way the organisation (or, more specifically, the people in the organisation) view itself and themselves, and their relationships with others, within everyday action. A good place to start on this would be the strategic-narrative resources on the Anecdote website, such as their whitepapers on narrative and change, anti-stories, and ‘making strategy stick’.
The other key point is that, again, it’s not something that came be done simply by executive diktat. It also is not amenable to ‘control’ in the classic management sense – much as Ed Catmull’s story indicates above. To make it work, we have to be in it for the long-haul: it’s a slow, subtle, very delicate process – and one that must be done with respect, or it can blow up spectacularly in our faces, potentially causing enormous damage to the organisation and its enterprise. But given the implications of that variant of Conway’s Law, it’s also the only viable option we have it we want and need to create deep structural change in the organisation that will actually last.
Hope that’s useful, anyway – over to you for comment, as usual?