First, though, an essential read to set the scene: Don Tapscott’s article “The World’s Unemployed Youth: Revolution In The Air?” (Note that that article was posted on the Huffington site on 4 June 2011 – more than two months before the riots began.)

An anti-client is someone who in some ways shares the same extended-enterprise as the organisation in scope for an enterprise-architecture – the same space as any of the organisation’s clients. But unlike a non-client, who is in the same conceptual ‘enterprise space’ as the organisation’s clients but has no interaction with the organisation, an anti-client will actively oppose, reject or object to the organisation’s presence in that extended-enterprise. Many businesses will have ‘inherent’ anti-clients: environmentalists in relation to oil-companies or miners, for example. And far too many organisations create their own anti-clients, converting previous good clients into active ‘enemies’ – anti-clients – through poor service, misleading contracts and all manner of other minor dishonesties and ‘game-plays’ that are all too common in the business context and elsewhere.

A key point here is that in ‘the good old days’ of broadcast one-to-many media, when companies all but controlled all access to the airwaves or the press, an organisation’s anti-clients had so little leverage that the organisations could usually afford to ignore them. Some careful PR would keep everything ‘on message’ without too much effort, or cost. But when the internet, SMS and other media allow many-to-many communication, or even many-to-one, organisations find themselves in a radically different game, where there’s no possibility of ‘controlling the message’ – and where even a single aggrieved anti-client can cause huge reputational and other damage with one well-placed viral video. Suddenly, an organisation’s anti-clients can have more power than the organisation itself – a very significant point…

A kurtosis-risk is a risk in which the losses that eventuate from the realisation of the risk exceed the total apparent gains made by ignoring the risk. (It’s sometimes called ‘long-tail risk’ because it’s a risk analogue of ‘long-tail‘ opportunities.) The important point about kurtosis is that it is usually a knowable or identifiable risk: it’s not an ‘unpredictable’, but one whose risk-pattern can be identified within the drivers for the statistical distribution of risk.

The risks from poor customer-service represent an all too common example of a kurtosis-risk: there’s no way to predict which incident will cause a massive blow-up, but we can predict that poor-quality customer-service will lead to a blow-up at some point. And we can also predict the statistical distribution of the scale of the risk in much the same way. And for organisations, one of the key drivers for the scale of risk – in other words, the positioning of the risk on the long-tail, and also the likely loss when the risk eventuates – is the separation between and effective power (or lack of it) of the agents of risk, such as disgruntled (ex)-customers. Separation between agents of risk is a risk-divisor: when separation is high, the effective risk is reduced, or pushed further down the long-tail (but never actually disappears). Separation is at its greatest – and hence risk is at its lowest – when the organisation controls one-to-many broadcast; or, to put it the other way round, the risk-multiplier increases, and the risk moves ‘up’ the long-tail, with increasing availability of many-to-many communication.

So, if we put the two together, anything that risks creating anti-clients, in a context where the media balance shifts from one-to-many (broadcast) to many-to-many (peer-to-peer), represents a context in which the risks are increasingly likely to be unacceptably high. They can no longer be ignored: they must be mitigated. And the only way in which the risk can mitigated is to seek out and pre-empt or resolve any context in which anti-clients could be created.

We can now apply this an enterprise at a very large scale: an entire socioeconomic system.

First, what are the anti-client risks? It doesn’t take much effort to identify that the socioeconomic model in place in Britain creates huge alienation, particularly amongst young males. Real youth unemployment is up above 50% in many inner-urban areas; those young men have have literally nothing to do, no apparent place in the society, no apparent means to gain social-status or even the resources that they need to live, and no apparent prospects or hope for change. It’s the same drivers that lead to the so-called ‘Arab Spring’ – in other words, that lead directly to active revolt against the state. Some people have expressed ‘amazement’ that the youth are ‘destroying their own community’ – but the key point here is that, to the youth themselves, they don’t feel that they belong in that community. It isn’t ‘their’ community: they feel rejected by it – and hence go off to create their own sub-community, in rejection of the ‘host’-community, and, in an all too literal sense, parasitic upon it.  (The crucial point to understand here is that they don’t feel that they have any other choice.) In effect, they have become anti-clients to the entire society in which – in theory only – they supposedly live.

In short, a perfcect recipe for a social explosion.

We then add in two other factors. One is the kurtosis-risk: the fragility of the economic system means that quite small acts of rebellion create disproportionately-large disruption. This is the basis of all asymmetric-warfare – including terrorism. The other factor is the availability of peer-to-peer communication, which enhances both the effective kurtosis-risk and the the cohesiveness of the ‘aliented’ group as a group.

Now add in yet another factor, namely rapidly increasing prices for essentials (the initial trigger for the Arab Spring).

And add in yet another factor, namely a government hell-bent on ‘cutting costs’ by shutting down any ‘unnecessary’ social programmes, especially in socially-stressed areas – at the same time as apparently providing massive subsidies to those already perceived as over-paid and undercontributing. (It doesn’t matter whether this is ‘true’ or not: the crucial factor is whether it is perceived as ‘true’.)

What we end up with is ‘an accident waiting to happen’: a context in which the ‘unexpected’ risks – already dangerously high – were being exacerbated in almost every possible way. And the actions that were known to be needed to reduce the risk were not carried out, on the grounds that they would ‘cost too much’. As Don Tapscott’s article makes all too clear, it may be a shock when it happens, but it should not have been a surprise. As Camila Batmaghelidjh put it in a much-praised opinion-piece in the London newspaper The Independent:

It costs money to care. But it also costs money to clear up riots, savagery and anti-social behaviour. I leave it to you to do the financial and moral sums.

The riots are a classic example of a kurtosis-risk: the cost-savings from cancelling those social-programmes were perhaps a few tens of millions at most, whereas the insurance-costs alone are already running into the billions. Ouch…

That’s what happens at a societal scale. Now bring it down a bit, to the scale of your own organisation. Look at the context architecturally, in much the same way as above: what do you have in the current architecture that risks creating anti-clients? What are the kurtosis-risks for your organisation, particularly around anti-client relationships. It can be an interestingly scary analysis… but one that’s well worth doing. Preferably right now, before the risks for your organisation get any greater?

The discussion went on for quite a long time, but it’s perhaps useful to record in some detail because it shows quite a variety of angles from which to tackle this specific type of objection.

This’ll be another long one, so I’ll place the usual ‘Read more…’ link here.

At the time of the session I previously blogged here, some of the key participants were almost beginning to grasp the notion that we needed to think about value in architecture somewhat wider than just ‘value=money’, and that we also needed to consider other business-models than solely the large US-style for-profit corporate. But then came another post to the thread that started the whole thing off all over again. The previous post (from someone else, not me) had listed four different categories of value: ‘economic value’, ‘quality value’, ‘utility value’ and ‘perceived value’. One of my regular US protagonists jumped in at this stage, with the aim of reducing everything back to ‘economic value’ – in other words, monetary profit:

Let’s remember that value is relative. I always think in terms of value to the classically defined constituents – the shareholders – but Tom is right that there are other values for other actors.

But quality value (quality of products and services) is a factor in economic value for a for-profit company. Listing it is therefore redundant, unless one is speaking of value from the perspective of a consumer rather than from the perspective of the organization.

If by “utility value” you are also speaking from the perspective of the firm, utility value is also redundant in the same way, because utility translates into business effectiveness, which eventually translates into economic value.

By “perceived value” perhaps you are speaking from the perspective of the consumer of the organization’s products or services. This also translates into economic value because customers who perceive value will return.

If speaking from the perspective of the firm, the above types of value are not redundant with respect to economic value only if one is trying to decompose the causes and effects that contribute to economic value. For a classical for-profit entity, economic value is the single end goal, and all other sources of value are merely a means to that end.

Which meant that the whole conversation had been dragged back to that one single special-case, the large US-style for-profit corporation. Perhaps unsurprisingly, I kind of lost my rag at that point:

We’ve gone through all of this before. After a brief respite (“Let’s remember that value is relative”) you’ve fallen straight back into the same fallacy again: “I always think in terms of value to the classically defined constituents – the shareholders”.

Please remember that this assertion of yours applies in precisely one business-context (the ‘stockholder is exclusive owner’ model) in precisely one country (the US) – not to all business in the whole world, which is the nominal context we’re dealing with here. Every other business-context recognises other interrelated forms of value; every other country recognises the dangers of placing money as the sole measure of value. (And in the US too – as in the classic Native American epithet that “only after the last tree has been cut down, only after the last river has been poisoned, only then will you realise that you cannot eat money”.)

It is also now definitively proven invalid in terms of behavioural science: a focus on money alone (as an abstract form of ‘extrinsic value’) will guarantee lower quality (by any metric) than a more balanced approach focussed on value as value (‘intrinsic value’). If you want to damage the effectiveness of the organisation, all you have to do is demand that everyone focus first and foremost on money. (See for example Daniel Pink’s book “Drive” for more detail on this.)

And even a few moments’ thought should show you that your reasoning could equally be applied to any other of that list of value-types, to assign that form of value as having the ultimate priority (for example, it is equally valid to say that monetary-value is a form of utility-value or perceived-value). But in reality it makes little sense anyway to try to reduce one type of value to another.

Remember too that the literal meaning of ‘economy’ is ‘the management of the household’. Money is a significant factor in managing a household, but it is by no means the only one, and as any parent would tell you, it is usually far from the most important. Describing money as ‘economic value’ is, bluntly, absurd: much like IT-centric misuse of the term ‘business-architecture’, it’s a ‘term-hijack‘ that arbitrarily constrains the scope to a very narrow subset of the issues, making it all but impossible to tackle the real scope as a whole. Try running a household on strict monetarist lines, based on the shareholder-value model: you’ll discover very quickly that it makes no sense at all. So why on earth do you think it would make any more sense in the much more complex economy of a business, a country or a world?

In essence, what you’re doing is circular-reasoning based on arbitrarily ‘privileging’ the assumption that money is more important than anything else. ‘Privilege’ literally means ‘arbitrary priority in law’ – and it does seem that the sole reason you keep asserting that money-for-shareholders is the only measure of value, is that this is what your country’s law says it is. But so what? Here in Britain we still have laws that require all adult men to practice archery on the village-green each Sunday (13th-century) and that make Christmas itself illegal (17th-century). They’re still technically in force because no-one’s bothered to repeal them, but time’s moved on a long way since then – and the exact same is true of the archaic notion of ‘shareholder-value’. And anyway, since when has ‘the law’ ever made consistent sense?

So yes, I’ll agree that the law about ‘shareholder-value’ is there, and we have to accept that. The fact that it makes no sense in practice is simply one of our many occupational hazards, that’s all. As architects, we need what works, right here, right now, in the real world. And key to that is the explicit acknowledgement that value is value – not money. Show the linkage to money, by all means, if you must; but if you want your architecture to work, don’t place money as the sole centre of your architecture.

Yet this didn’t get through either. Even though the protagonist made various attempts at clarification, it was clear that he was still forcing us to keep running round in the exact same circles:

I don’t feel that money is more important than anything else. I was careful to stress “for-profit company”.

I just think that “since the law is there” as you say, the place to fight this is in the political realm, not the BA realm. After all, a BA is merely an analyst of sorts and is not a decision-maker. If you try to force your political view on the decision-makers you will be without a job pretty quickly!

In other words, once again, he’d missed the point, and I said so – perhaps a bit more forcefully this time. So once again, I tried to find yet another way to explain this same distinction between the ‘theory’ defined by external law and politics, versus what we actually need to do in practice to get an enterprise-architecture to work:

I think we’re agreed that current corporate law is inane, if not insane, in every sense of the word. It doesn’t work; more to the point, it cannot work, especially in the longer term. The proof’s been there for decades, if not centuries: we all know this.

But yes, the law is the law, and yes, we are required to deliver against it, whether we like it or not. And also yes, to change it is a political matter, not an architectural one, so that part of it is outside our scope.

Yet the law, and its underlying assumptions, simply do not work, either in theory or in practice. Again, we know this. So we also know that if we were to be so foolish as to design our architecture, internally, on the same assumptions as the ‘external’ law, we will all but guarantee failure.

What we need to do instead is design our architecture internally on value-as-value, on enterprise-as-enterprise. We then, separately, show how this links into monetary return. That transform is (relatively) trivial, and (relatively) easy to document. What we don’t do is use monetarist notions within the architecture, because we lose so much key information and miss so many key dependencies that there is no possibility that it can be effective.

It’s exactly the same problem as IT-centrism – trying to force-fit everything into a single domain via a single filter. It does not work – we know this. So please don’t replicate the same problem elsewhere!

This isn’t politics: it’s pragmatics. It’s strictly about what works, and within architecture: nothing more than that. If you can’t see the difference, well… dunno what else to say, really.

By this point I’d thought we were finally getting somewhere. But no, evidently not, because someone else – from the British contingent this time – came back to the same red-herring about politics:

I’m sorry but I have to agree with [my protagonist] on all this.

Business architects, in the majority, feed from the corporate watering hole so you don’t bite the hand that feeds you or poison the stream; whether it’s right or wrong you play the game or, to translate his point into UK speak, it’s “P45 time”. ['Pink slip' is the equivalent US term.]

If one was having discussions about value in this way in a corporate workshop I can think or several senior managers in my past who would be down to HR seeking your severance. Politics and business architecture don’t mix in my view. I have issues with offshoring and outsourcing of UK jobs but I have learned, to my cost, to keep my trap shut!

I am not questioning the substance of your views which may well be right philosophically, I do have sympathy with them, but this is business change land where we continually try to justify our existence with hard value to the business we work for, and stuff like this doesn’t really help the cause.

So, once again, back I went round the same old loop:

It’s not about politics.

It’s about what we model, and how we model.

If you don’t model values as values, you will miss key information and key dependencies. That means that your architecture will not work.

If you don’t model enterprise as enterprise, including all of its values, you will get your own equivalent of ‘United Breaks Guitars‘, time after time after time – but you will have no means to identify why this is happening, because you will have excluded the dependencies from your models.

The transforms to show how each of the values impacts on ‘shareholder-value’ and the like are relatively trivial. But they are not reversible, because they are complex non-linear derivatives. You therefore cannot use monetary metrics alone to guide the architecture itself.

This is the basis of long-established frameworks such as GRI and Balanced Scorecard: it’s nothing new, nothing special, nothing ‘political’. It’s just what works.

Monetary-metrics are crucial output-factors for an architecture. But they are not merely useless but actively misleading within an architecture.

If you can’t get this blunt fact over to your execs, go straight down to the HR department and ask for your own P45, because you’re wasting your time as a business-architect in that company.

Simple as that, really.

More comments ensued, still indicating that almost all of the active participants were still hung up on the notion that starting from values rather than money was somehow ‘anti-capitalist politics’ – which is nothing to do with it at all. Hence I spent some time giving a detailed example, describing some work I’d done about a year back with a C-level exec for a very large semi-public corporation. (I won’t name the corporation on a public site, but please contact me direct if you really need to know.) The example was as follows.

During what was scheduled to be a one-hour discussion, the exec asked me to look through their supposedly ‘final’ presentation on their intended strategy and plan for a fundamental change to the structure of the corporation. But it turned out that they’d built their model and entire plan solely on monetary metrics, restricting the strategy solely to those items that could be measured in monetary terms – such as facilities, machines, FTEs, assumed productivity, financial ROI and so on. Within less than three slides I could see that there were already several fundamental flaws in their reasoning, because they had only used monetary metrics. So I got him to stop, and instead we just quickly flicked through each slide of the rest of the presentation, so that we could identify the nominal assumptions in each case.

We then went back to the start. I first walked him through the basic principles I’ve described here – that most transforms to monetary metrics are non-reversible, and that we can therefore use those metrics only where there is a clear linear relationship, such as for purchase-cost or consumables. For everything else, we have to start from the respective values, and identify the factors in the respective non-linear, non-reversible transforms to the final monetary values.

The whole thing took just slightly over the hour. By the end of that time, it was obvious not just to me but to him as well that their strategy, as described, was a guaranteed recipe for disaster. More to the point, we had summarised every key hidden-assumption, why that assumption was invalid, and how to restructure the strategy in a way that would actually work. We then parted, with many thanks from him.

And ‘thanks’ was all I got from the exercise, of course – no further work, although that’d been the nominal aim of the meeting. (Yep, you could call that a ‘P45 moment’, I guess – though I think it was more about avoiding potential embarrassment at the scale of the blunder implied in their initial strategy.) But I did have the satisfaction of seeing in the newspaper some weeks later that they’d announced a new strategy, in essence based on the rework that we’d done in that one hour.

The savings to the corporation implied by that one-hour meeting? More than one billion dollars. That’s the minimum in direct costs that the original strategy would have lost them – not to mention huge indirect losses from industrial action, market rebellion and the like, probably also in the billions of dollars.

So on the LinkedIn thread, I reiterated the point that whilst politics and so on are indeed extremely important as factors whose dependencies and interrelationships need to be modelled in the values-analysis, politics should play no part in the analysis itself. It’s input, not process. So if, for example, the only outcome of concern is short-term profit or shareholder-value, that’s fine (probably): but don’t try to model it solely via monetary metrics, because it doesn’t work. In that example, it would have been literally billions of dollars of ‘doesn’t work’.

And yes, I’ll admit that, like everyone else, I do have specific political views: but in essence, those views – such as they are – come from the architecture-methodology, not the other way round. In effect, the sole relevance of politics to the organisation is as part of the intersecting mesh of values that impact on the business of the organisation: anything else is out of scope for EA or BA. Professional discipline requires that, as far practicable, we keep our own views outside of the mesh.

At the end of the example, I summarised:

So it’s not about politics. It’s just about process – the process of doing architecture in a way that actually works. That’s all that I’m concerned with here. But to make sense of this, you need to be able to separate the politics from the process: and I can’t do that for you. That’s a choice you have to make for yourself.

But even that didn’t seem to sink in, as the next response from the British protagonist demonstrated all too well:

This is why architecture fails in so many places because this type of stuff in a deep debate is too far away from the price of fish. Senior guys just get bored with it and see it as waste they then ignore you/us. I have seen this so many times where people just get sidelined because they “think too much” I saw one guy loose his job and retired early because he got too political – the guy was seen as an irritating left wing nuisance and they let him go. – No reflection on yourself naturally so please don’t take offence because it is not intended.

Ignorant yes but that is the way of the world; as I said before Keep it Sweet and Simple; because that is what they pay you for its our core skill.

We all might not agree with this but at the end of the day engaging in endless academic debate and they will not appreciate it; it ain’t commercial.

I think we both get what you say but don’t see it as a pragmatic commercial approach.

I’ll admit that by this point I was starting to get more than a little frustrated, and hence started again with a yet another different approach to the whole mess, trying to find any way to break free from the ‘politics’ red-herring:

One more try…

This is a business-architecture list, so let’s do this in strict business-architecture terms.

Step 1: Go to the Business Model Designer website at http://bmdesigner.com/, which provides an online, interactive version of Alex Osterwalder et al’s excellent ‘Business Model Canvas’.

Step 2: If you’re not familiar with the Business Model Canvas, review the embedded presentation on the site. Or got to http://www.businessmodelgeneration.com/ to download the very generous 72-page preview.

Step 3: Note that the core of the model is the value-proposition(s) that the organisation presents to the selected group(s) of stakeholders. In a for-profit business, this connects with revenue-generation, but note that value-proposition and revenue-generation are modelled separately in the Canvas. Note also that it is called ‘value-proposition’ for clear, explicit reasons, and that it is not the same as the monetary (or equivalent) transaction that may accompany it.

Step 4: Note that much the same applies on the ‘cost’ side of the model, with separate but interlinked modelling of costs from the ‘cost centres’ themselves.

Step 5: Define your business-model(s) using the Canvas.

Step 6: Now step back a bit, and widen the scope from from money-only transactions. For example, you must present a value-proposition to a prospect before any transactions can take place (hence separate modelling of channels and customer-relations in the Canvas). You also have value-propositions that link to many groups of non-customers, in government, the community, the future (usually simplified as ‘environment’), ex-customers, anti-clients and so on. Almost by definition, these value-propositions do not transit through the transaction-economy but are equally critical to the success of the organisation. (Instead, they transit through various combinations of the attention/respect and reputation/trust economies, on which the transaction/money economy ultimately depends.)

There is no difference in the basic approach between Step 5 and Step 6. The only difference in including Step 6 is that we take care also to model value-propositions in the other non-money dimensions of the market context.

This is critical, because ultimately everyone in the extended-enterprise is a ‘customer’ of the organisation, whether they buy anything or not. The market dependencies can be summarised as reputation -> trust -> respect -> attention -> transaction -> money. So unless we do model all of this, we would have no means to understand where our monetary profit is actually coming from – or where it could vanish without apparent warning.

If you want to explain that you’ve built an architecture that has no means to detect how the company’s market could vanish overnight, be my guest.

If you want to build something that can identify what’s going on both inside and outside the organisation, model values as values, and enterprise as enterprise, much as described above.

At this point there seemed to be the glimmering of an awareness that perhaps I wasn’t describing something ‘political’, but that it really was about business and business-architecture. To hammer home this home, I returned to the previous complaint that this was somehow ‘academic’ and ‘not commercial’:

Look back at the details of the ‘United Breaks Guitars’ incident. That had an immediate direct cost (damage-control PR etc) to United that would have been up in the low millions, and a short-term indirect cost (lost-revenue, follow-on damage-control) probably in the low tens of millions. There was also a significant but arguable short- to mid-term hit on shareholder-value, and incalculable longer-term to brand, trust-relationships etc, but let’s ignore those for the moment: just stick with the ‘low tens of millions’ figure. Say $20m, for a nice round figure.

Would you be willing to say to an exec that discussion of the factors behind a $20m+ loss is “academic” and irrelevant?

And remember that that was for just one incident. Yet there could be many more such incidents. Which could come from anywhere, without warning – because the factors that track it are not yet included in your architecture. Would you describe that as “academic” too? I would hope not…

In academic terms, this is known as ‘kurtosis risk‘ or ‘Taleb distribution‘, in which a high probability of many small gains (such as, in this example, by not paying damage-claims) is matched by a low probability of very high loss (‘United Breaks Guitars’) which exceeds all of the accumulated small gains. A business-model with high kurtosis-risk is very risky indeed, especially in the longer term. But in this case – as in many others – the only way to identify and model the risk is by modelling values as values – not solely as monetary metrics.

May look academic at first, but it sure ain’t academic in practice.

And at that point – at long, long last – it seemed as if they finally got the point.

Watch this space?