Why vision? Whose vision? What do we mean by ‘enterprise vision’, anyway? And who’s responsible for it? – who should create it?
This enquiry arose from a great multi-way Twitter-conversation following my previous post ‘Yes and No‘:
- tetradian: [post] Yes and no: a question of commitment http://bit.ly/fJUqcA #entarch #culture #responsibility
- MartinHowitt: @tetradian if an explicit vision does not exist for an enterprise, is it the #entarch responsibility to facilitate one?
- tetradian: @MartinHowitt: “if an explicit vision does not exist for enterprise, is it #entarch responsibility to facilitate one?” – my opinion is Yes // enterprise-vision is also anchor for ISO9000-type quality-system – it’s essential/fundamental part of the architecture
- NGA_Anita: @tetradian @martinhowitt an enterprise vision is first a business responsibility, #entarch should never take this on.
- tetradian: @NGA_Anita yes, vision is biz-responsibility: @MartinHowitt and I both said ‘facilitate’, not ‘create’ – we need it to exist for #entarch // e “#entarch should never take this on”: if you mean IT-centric ‘entarch’, probably agree – it’s a biz concern (real-entarch)
- erikproper: @tetradian Isn’t that explicit vision part of the EA process …. step A: Enterprise Vision ….
- tetradian: @erikproper TOGAF Step A ‘Architecture Vision’ describes ‘future state’ for architecture – transient condition, not permanent anchor
- CarlChilley: @MartinHowitt @tetradian re vision: no vision = no ent just org with no purpose => no #entarch just collection of stuff with no context
- MartinHowitt: @CarlChilley @tetradian spot on. We need stuff to work with! // OTOH, “vision” can take many forms. It doesn’t have to be in a “statement”
- kvistgaard: @CarlChilley there is always a ‘context’, the missing thing would rather be ‘cohesion’ (I agree with the rest of the statement)
- CarlChilley: @kvistgaard cohesion and coherence both necessary even (or especially) when context is obscured
- CarlChilley: @MartinHowitt @tetradian vision does take many forms-observation spot on. Vision as behaviour, vision by example subliminal & often ignored
- MartinHowitt: @CarlChilley @tetradian best #entarch thing I ever did is chat to a senior mgr, write his vision down & hand it back. Smiles all round
- CarlChilley: @MartinHowitt @tetradian capture and present verbal #entarch always works with reasonable people. But egos can often complicate the simple
- seabird20: @erikproper @tetradian we must be very careful we dont “do architecture” for its own sake.
- tetradian: @CarlChilley @MartinHowitt @seabird20 re vision: needs longer reply, will write blog-post on this later today
Hence this post.
Those questions about vision are right at the core of enterprise-architecture: yet they’re so often misunderstood that we really need to address them right from the start of any enterprise-architecture work. If we don’t tackle them early, it’s likely that a lot of effort will be wasted, and a lot of rework done at a later stage. It matters.
So: a bit more detail…
The role of the vision is to act as a guide for decision-making. When there are no explicit instructions or guidelines, or wherever there is uncertainty in a context to which the existing rules and guidelines do not seem to apply, the vision provides a definite point at which to aim.
Whenever a decision is in doubt, the vision provides the ultimate ‘Why’.
In enterprise-architecture, the vision ‘belongs’ to the commissioning organisation – in other words, it acts as a guide for decision-making within that organisation.
However, the vision will usually describe a context wider than that of the ‘owning’ organisation (see slidedeck ‘What is an enterprise?‘ for more details on the distinction between ‘organisation’ and ‘enterprise’). The broader scope of this ‘extended-enterprise’ implies a need to summarise not only the direct concerns of the owning-organisation, but any ‘external’ business-drivers for that organisation. For an IT-oriented enterprise-architecture, where the ‘owning organisation’ is typically an IT-unit or IT-department, the content for that organisation is the physical and virtual IT-infrastructure; but the scope of the vision would need to include the applications that run on that infrastructure, the data manipulated by those applications, and the business-usage of that data as business-information. (This is the scope of ‘Architecture Vision’ described in the TOGAF-9 enterprise-architecture framework.) For a business-oriented enterprise-architecture (typically starting at the level at which TOGAF ends), the vision would need to encompass the concerns of the supply-chain, the customer-base and other stakeholders – government, investors, community, anti-clients etc – beyond the basic supply/transaction layer. In essence, if the owning-organisation is thought of as ‘Self’, the vision anchors decision-making not only within the Self, but also in relation to anything ‘not-Self’ that might impact ‘Self’.
What kind of vision?
There are at least three types of ‘vision’ that we come across in enterprise-architectures:
- a description, often in the form of a diagram, of some desired ‘future state‘ of a system or structure, as delivered by a project or portfolio of change-projects
- a ‘positioning statement‘ about where the organisations intends to be, often relative to competitors or market, at some specified point in the future
- a ‘permanent anchor‘ that summarises the concerns and drivers for all players in the extended-enterprise in which the owning-organisation chooses to plays its part
The ‘future state’ form of vision is constrained to the scope of a project: it is valid only for the life of the respective project, and often only for the earlier stages of that project. This is the form of vision described in TOGAF-9: a different ‘future state’ vision should be developed for each architecture-iteration.
The ‘positioning statement’ form of vision is typically developed as an artefact of a marketing or motivation exercise. This is the form of vision described in conjunction with the BRG/OMG Business Motivation Model. In effect, it is a more emotive form of ‘future state’ vision. It is valid for that purpose, for the life of the respective marketing-campaign etc, but is usually self-focussed and often circularly self-relative, and should not be used as or mistaken for the ‘permanent anchor’ form. (One of the reasons we should not use it as an enterprise-anchor is that it literally excludes others – giving them an explicit absence of reason to connect with the organisation. In fact, in the infamous form that focusses only on the organisation’s profit or ‘shareholder value’, it gives customers and others a very good reason to not engage with the organisation… not wise!)
The ‘permanent anchor’ form provides a summary of the overall direction of the whole enterprise in which the organisation is engaged. This is the form required for whole-enterprise architecture, and also as the anchor for ISO9000-style whole-of-organisation quality-systems. In principle, it should never change; more to the point, if it does change, what it describes is no longer the same enterprise. A workable vision of this type is always brief – no more than half a dozen words – yet emotive enough ‘to get us out of bed in the morning’. It always describes the drivers for the overall context, that engages everyone in this shared-enterprise – not solely the drivers for the business itself. To me, it has three distinct components:
- something that identifies for the content or focus for this enterprise
- some kind of action on that content or focus
- a qualifier that validates and bridges between content and action
These components may occur in any order, but all of them need to be present in a vision-descriptor. The vision for the TED conferences, “ideas worth spreading”, illustrates this well: it’s clear, succinct, emotive, and conforms exactly to that structure: “ideas” [focus] “worth” [qualifier] “spreading” [action]. A bit more detail on this, from my book Mapping the Enterprise:
Note too that none of these items describe the organisation at such – but do describe the focus, the area of action, and the key value-metrics which define the meaning of ’success’. That’s what we need to look for at this stage.
The content part of the descriptor should be a noun that identifies and summarises the ‘things’ of the enterprise as a whole. These will usually not be products or services as such – because that’s what the players in the enterprise deliver to each other in order to create value in the enterprise. Instead, it’s more about what those products or services act on, and are of interest to every player in the enterprise.
The action part should be a verb, summarising the overall activity of the enterprise. If at all possible it should be explicit and distinct, rather than a generic such as ‘supporting’ or ‘making’ or ‘creating’ – though such generics can be useful to give something to work with whilst searching for a more distinctive vision.
The qualifier part should be an adjective or suchlike that is both emotive and implies some value or valuation – for example, the adjective “worth” in the TED vision of “ideas worth spreading” would imply some means and value-criteria to identify which ideas are worth spreading, and which are not.
To identify the vision, look around both within and outside the organisation for anything that engages or excites people in their work and in their relations with others in the shared-enterprise. And excitement or engagement is the key here – though note that it can often take somewhat strange forms.
The vision implies a narrative, a story, the enterprise as story. It implies the values of the organisation in relation to the enterprise, and thus invites in (rather than drives away, as the ‘marketing-puff’ form so often does). It indicates what ‘effectiveness‘ actually means within the organisation and enterprise. It implies metrics against which the organisation (the enterprise as a whole, actually) expects to be monitored and measured.
And so on, and so on: I could probably ramble on about this for hours, but probably shouldn’t…
But I hope this gives something useful to start with, anyway?