Value network: a non-value-add critique of the value chain concept

2/18/2007 This article and associated debate is receiving a lot of hits. It's essential for the newcomer at least to read my IT Value Chain (excerpted from my book).

Also of interest may be:

The two faces of demand management

IT value chain from Brazil

-ctb

==========================================================

Recently I've become aware of a value chain critique called "value network." As stated on Wikipedia,

"The value networks model challenges the traditional notion of a value chain. Historically we have been in an industrial age, focused on a linear value model, and have recently begun to switch to a new business style in which there are a web of different resources that work together to create value."

For reasons that will become clear to all of you soon enough, I need to get on the record here. For the purposes of modeling enterprise IT service management, the value network concept adds no value to Porter's classic value chain, and, by obscuring the all-important distinction between primary and supporting processes, actually does harm. In fact, it appears to be a reaction to a caricature of Porter, not a fair reading.

I have Porter's seminal Competitive Advantage here on my desk. In his discussion of the value chain, in no place does he state that each step of the value chain is always sourced by one player, nor does he insist that the value chain be restricted to one firm. In fact, he is clear that the issue is complex, and that firms may achieve their value chain by increasing their scope or by partnering with others. 

His key insight is that - regardless of sourcing - value is derived from a logical sequence of activities. Supplier sourcing precedes manufacturing, which precedes distribution, and so on. The fact that any one of these sub-processes may have one to many parties fulfilling them is not relevant. That is an implementation issue.

What attracted me to the Value Chain concept is the distinction between primary and supporting processes. We struggle with this mightily in enterprise IT: architects versus developers; security staff versus operations. A conceptual framework to help us manage these contradictions is essential, and the Value Chain is the most useful I've seen: it's concise, accurate, and business people relate to it (because most of them have seen it in their MBA programs).

In enterprise IT, a Value Network is an instantiation of a Value Chain. They are not contradictory. Posing the Value Network as an "improvement" on the concept of Value Chain indicates a lack of understanding of the concept of abstraction, and ultimately is a naive critique. Value Chains are abstractions of Value Networks. They are how we understand and analyze Value Networks. To lose the abstraction is to dive back down into the muddy details that the Value Chain theory has started to help us clarify.

It is inevitable, of course, that an academic of the stature of Michael Porter should have attracted his critics. That is the nature of academia.

Yours,

Charlie

Comments

[Kimberly]

I have to admit a chuckle as I read your state-of-affairs on the Value Chain. It brought me back to the vigorous debates in the 1990s.

This critique is neither new nor wrong. The fact that it involves Porter is also not new. Do you still adhere to Porter's 5 forces model? I bet Mr. Porter wishes he could take back that one.

Top business schools teach three models: Value Chain, Value Network and Value Shop. Even Porter's own Harvard. Information-based economics revealed weaknesses in the VC and new models were developed. This conversation began almost 10 years ago and is well established dogma. Certainly open to improvement but both industry and academia are way past the VC model. I noticed both HP and IBM, for example, adopting VN and VS frameworks in their services organizations.

Heck, one of the key inspirations for the Value Network came from down the street from Porter's office at MIT Sloan and their work on System Dynamics.

A Value Network is not a collection of Value Chains. Need proof, how would describe Wikipedia in a Value Chain?

Regards,
KH

[alphasong]

All such models are neither right nor wrong, they are only more or less useful. The questions are,

What is the most appropriate model to apply to enterprise IT in a primarily captive business model, but with diverse sourcing?

How can we understand its perpetual contradictions and struggles to achieve governance?

How can we break down the functional silos between the governance, development, and operations areas?

I continue to believe that a simple value chain model is appropriate and useful for enterprise IT service management, I am successfully applying such in my day to day work, and I am not convinced that either value network or value shop add any value here.

I have no opinion on their utility for other business purposes.

-Charlie

[Kimberly]

You are correct. All decisions are based on models, and all models are wrong. These statements are deeply counterintuitive and requires humility about the limitations of our knowledge.

That is precisely how these new models began. To help people open up to a new perspective, a new model, and change
deeply entrenched behaviors.

We fail our bosses, clients and students when all we do is facilitate the old organizational change recipes. We may only succeed in replacing one dogma with another, helping people develop the critical thinking skills to continually challenge their own models, to uncover their own biases.

When Porter wrote Competitive Strategy (1985), business was a different place. A service business operated much like a manufacturing business with assembly lines of suppliers, manufacturers, distributors and retailers.

(Although you may find it interesting to note, it was large manufacturers like Dow and GM that were the first to adopt VN models in their service lines (e.g. GM financial arms) )

It was precisely IT services which caused many to go back and rethink their models. Even Porter. I'd recommend spending a bit of reading time with the bloke who started this, Jay Forrester of MIT.

Forrester developed a well-known simulator, called the Beer Game, which experientially demonstrates the limitations of linear VC thinking within the supply chain. Several vendors have adapted it to ITSM. They are not as rigorous as the Beer Game, but instructive nonetheless.

It is difficult to pick up a quality modern business book on governance, organizations, process or policy without feeling Forrestor's influence.

Cheers,
KH

[alphasong]

OK - all interesting stuff and now I know more than I did - thanks. But fundamentally I'm not convinced. Some Google search results:

"Value Chain" : 1.3 million
"Value Network" : 194,000
"Value Shop" : 87,800

Value Chain has an industry association, conferences, and even a markup language (Lord knows what for...). The alternative terms have nowhere near the mindshare.

Actually I have come across Forrester before. Can't argue with the man's credentials of course, but I think he was addressing different issues. Cybernetics and systems theory I would see as a set of tools to help us understand chaotic value chain dynamics. But the basic linear abstractions still hold. In the IT world, we:

1. Identify and validate customer demand
2. Design and build solutions
3. Support services

along with a set of supporting activities that are constantly "getting in the way" of the people doing the primary value chain. When I've put this analysis in front of people, light bulbs go off. The approach has utility.

Value network, on the other hand, just looks like a messy functional decomposition with forces going every which way. I've built dozens of those maps (haven't called them value networks, but the modeling is the same), and they always suffer from the question of *priority*. (This is based on the models I'm seeing cited off the Wikipedia articles, e.g. http://www.vernaallee.com/value_networks/ Understanding_Value_Networks.html )

Value Shop just seems to be a recasting of the Value Chain, substituting different activities for the manufacturing stuff. I always assumed that the specific activities in the primary value chain could be altered; that is the basis of the model I've done.

When and where did the "vigorous debates" occur? What leads you to say that HP and IBM are using VN/VS frameworks? I would still assume that they have a fundamental "book to close" and "order to cash" business imperative.

Regards,

Charlie

[Kimberly]

Your "proof by majority" is a bit illogical. McDonalds make more hamburgers than my grandmother, but it doesn't mean it tastes better.

In the early 1990s, Philip Evans and Tom Wurster realized there was something more to the Internet than media convergence. They began to trot out their ideas at conferences. They first published in the 1997 HBR article, "Strategy in the New Economics of Information."

Other notable papers (to name a few) on the deconstruction of the Value Chain include (notice how the titles get bolder), Frantz and Hout's 1993 "The New Vertical Integration", Carl Stern's 1998 "The Deconstruction of Value Chains", and Evan's 1998 "How Deconstruction Drives De-Averaging."

BCG was at the forefront for predictable reasons. Business schools caught on shortly thereafter. The conversation now is taking place at entirely new levels: Benkler's The Wealth of Networks or Chesbrough's Open Business Models, to name but a few books.

On the topic of Forrester, please do more than a casual glance. His work on System Dynamics shows us, in a quantifiable and repeatable way, that linear models (when applied to non-linear problems) are destined to fail. Usually in the same predictable way. As a introductory teaching tool, linear models are fine. Sort of like teaching Newtonian physics in high school. These tools are used to understand the oil problem (Hubbert's Peak), Economics, Supply Chain and, yes, IT organizations.

You've hit on a key point. Value networks are indeed messy. Things may appear clean and orderly when depicted in linear models. But that doesn't eliminate the non-linearities from real-life. That is a clue that we need something better.

Which leads to the real beauty of value networks. Once the organization begins to think in non-linear terms, they are one step away from applying modeling and simulation tools to measure improvement initiatives.

How's that for neat? The ability to use computer simulations to empirically prove/disprove the benefits of organizational improvement efforts.

All this talk about governance, CMDBs and such, move away from the subjective to the simulated. But we cannot move forward, it won't male sense, when the organization is limited to thinking in linear models like the Value Chain.

[Kimberly]

Charlie -

I realize I'm being a bit unfair. Moving from linear to non-linear thinking is not easy nor intuitive. Let's start over with a simple thought problem that has nothing to do with IT.

Some years back, Pres. Clinton announced he was sending $100M to poor Texans suffering from a heat wave so that they may purchase air conditioners.* What is wrong with this policy?

Clinton's linear solution is neat, orderly, logical and deeply flawed.

1 - Air-conditioner's don't make heat go away. They just move it from indoors to outdoors; worsening the heat island around the city, creating a need for more air-conditioners. A vicious cycle.

2 - The solution prevents people from looking at alternative, more enduring solutions. Such as light-coloured paving or non-black roof-tops. Or money for better insulation or controls for better vehicle efficiency.

Every time we have a problem that keeps getting worse no matter what we do about it, we are probably, in the most well meaning way, doing something that makes the problem worse.

This is what happens when we apply linear models/solutions to non-linear challenges. Can you see this pattern being played out every day in IT organizations?

Now try describing these IT non-linearities, particularly the intangible ones, in a value chain. The fact that they are primary or secondary does not really matter. In non-linear systems, effect is rarely proportional. Small things, over time, have huge effect. But that is another discussion.

I admit this may, at first glance, seem like an over-complication. You are a bright bloke. Check it out, spend like time with it. Look at the work being done today with System Dynamics. VNs may not do it for you, but you may find yourself deeply unsatisfied with Porter's model.

*credit to Donella Meadows of Dartmouth for pointing out this real-life example

[alphasong]

"Proof by majority" would be entirely illogical in a hard science or purely quantitative endeavor. But we are dealing in the slippery, subjective world of what some call "upper ontologies" : abstractions such as function, process, capability, entity, and so forth. In this inherently consensus-based world, mindshare does matter - there are endless numbers of thoughtful people (academics and others) coming up with their own idiosyncratic versions of intellectually "pure" frameworks. McDonald's very ubiquity *is* part of the argument. I need concepts I can sell.

One of my requirements is accessibility and familiarity, so a Google search *is* a useful data point. My audiences, in assessing the validity of *my* arguments, are going to use Google. Value shop in particular is pretty dismal, as a large portion of its hits don't even refer to the term as we are discussing it.

I'm a fairly hardened skeptic at times, and we both know that management philosophy has fads and fashions like anything else. I do intend to review the sources you cite. I think our fundamental disagreement might center on whether value network is a true paradigm shift from value chain, or whether it is merely a peer form of analysis. A paradigm shift in the strict Kuhnian sense (which you imply by your reference to Newton) would mean that the value network subsumes and incorporates *all* of the explanatory power of value chains. I think that it quite clearly does not; it does not inherently distinguish between primary and supporting processes, nor does it give us a repeatable series of events, at least in the admittedly limited readings I have done to date.

I rather view the value chain/value network dichotomy as simply the age old distinction between function and process. Functions are steady state activities that exchange services dynamically (value networks). Processes are repeatable event driven sequences that add value for some end stakeholder (value chains). Linear processes always are platformed on steady state functions (aka services, aka capablities.) Both types of analysis are important tools in the overall enterprise architecture toolkit.

In terms of IT governance: the organizations I have seen suffer from a lack of value chain view. We are caught up in endless attempts to govern something, but we are not sure what. I think it is largely a question of maturity. Most IT organizations I am famliar with have just started to successfully move out of a very primitive functional silo view and get to what we might call Hammer 1.0: the existence of a few processes that truly cross functional boundaries. But overall sense of a value chain is not even present; all we have are the feuding functions (development versus operations; applications versus infrastructure.) All that a functional analysis would do (and I have done several) is give us a bunch of boxes and lines, showing the various influences bouncing around.

Once we have inculcated the sense of a common value chain purpose throughout the enterprise IT capability, then we may need to improve the processes through understanding their non-deterministic, chaotic dynamics, and a value network might be part of that analysis. But it would not replace the value chain, as so much of the VN literature I am seeing implies. It's just another tool in the box.

Regards,

Charlie

[alphasong]

I appreciate the concept of non-linearity; in fact a few years ago I gave my stepfather the book _Why Things Bite Back_. Of course there is much in the large, complex IT capability that is non-linear. The CMDB in particular can be seen as the repository for attempting to manage IT non-linearity. The web of IT systems in the data center could be seen as a value network; most certainly NOT a value chain.

But in this view, the value network concept is subsidiary to the value chain. The value chain includes both the primary process of adding a new service to the value network, while the supporting processes attempt to govern the knock-on, chaotic, non-linear effects (such as capacity overlimit, security breaches, project overruns, major availability outages, and so forth).

I think that it is critical to distinguish primary and secondary so that we can achieve a balance of dynamic tension. Otherwise, the IT shop falls into one of two very well known pits: the "cowboy" shop at the mercy of nonlinearity, or bureacratic stasis where risk management becomes the entire purpose of the IT organization, and innovation slows to a crawl.

One of the insights that value chain thinking gave me in writing my book was the idea that the CMDB needed to be populated by the primary value chain. The trouble with IT governance today is that we are attempting to govern the non-linear chaos through isolated functional silos such as architecture, security, and risk management groups, each of which need the same base data. The practice in too many shops is to permit those groups to go out independently and attempt to re-harvest much the same information, a task for which they are ill suited in terms of process positioning and resources. If instead we insist that a primary part of the IT value chain is documenting all the IT dependencies in the CMDB, then we have some chance of managing the non-linearity.

I have seen that core IT value chain in operation in organizations with a total $5 billion IT footprint. While it has chaotic dynamics, those dynamics are not caused by the value chain representation - the map is accurate. Is it preventing us from alternative ways of thinking? I don't know, because it itself has not been widely disseminated; B-school MIS texts don't talk about it, instead (from what I've seen) emphasizing a functional mindset with all of its problems.

(you can see from my blog that there are a number of people wrestling with IT value chain; it certainly is not a concept unique to myself.)

Regards,

Charlie

[Kimberly]

In this $5B environment, is the CIO's initials RS? If so, I'm intimately familiar with the environment. Don't let the press releases fool you, they are having serious problems. There have been quite a few departures/firings since "2006". You'll notice RS now adopting terms like "complexity" and "systemic approach."

He confuses the meaning at times. Sometimes he means combinatorial complexity; making decisions in the face many components or possibilities. Othertimes he is refering to dynamic complexity; the counterintuitive behaviour arising from many interacting agents. A CMDB helps with the first, not with the second.

But it is obvious I've not moved you. If the logic is now based on the ease of "selling" ideas, then it is time for me to move on. These concepts are not easy to sell. Humans are not wired to be non-linear thinkers. Precisely why these problems persist.

Good luck with your value chain approach.

[alphasong]

No - I don't live in Detroit - it's been a combination of several, still very large IT orgs at which I've had consecutive experience across 4 industry verticals.

Please don't misunderstand - this has been a valuable discussion and I appreciate your perspective. I'll be doing further reading. Nothing I have ever written or said should be construed as saying that IT can be reduced to linear models - clearly, it cannot. It is merely a question of utility in the present context.

As we agreed earlier, *all* models are wrong. I just am objecting to a line of thinking I have recently seen in a draft version of a publication that promises to be highly influential. These authors took a hard line that the value chain model has nothing to offer our understanding of IT service management, and the value network concept was clearly superior. I will continue to object to that line of thinking. It smacks of B-school faddishness and one-upsmanship elitism. I have far too much sad experience in such polemical terrain, and have witnessed the decadence of fashion too many times to quickly board any such bandwagon.

I might be convinced that the approaches are complementary.

Re: "selling ideas." I will leave you with two quotes:

- Politics is the art of the possible
- In politics, timing is everything

Those without a core understanding of these principles, will never be change agents. They will remain in their ivory towers.

Yours,

Charlie

[alphasong]

P.S. I think a CMDB can also help with dynamic complexity. Counterintuitive, nonlinear behavior emerges from poorly understood dependencies. It is admittedly a more difficult problem. But I can envision a sophisticated IT analyst, or simulation program, parsing a sufficiently robust CMDB's dependency maps and identifying potentials for dynamic and unexpected interactions in the event of some change - I do not think there is any fundamental issue preventing that.

[stratman]

Kimberly – Good stuff. Nice insight. I’m a partner at a M/B/B strat firm and would love to chat further. We deal with these topics frequently am always looking for big thinkers.

These models are foundational skills for strategy consultants and would expect them to articulate them in this fashion:

Value Chains: focus on transforming inputs into outputs. They are most useful in the manufacturing and distribution of tangible goods. Based on a linear flow where tasks are accomplished serially with precedence constraints (interdependencies are sequential). Tasks tend to be repetitive. Good for standardized products and support the generic strategies of low-cost producers.

Value Shop: focus on customer problems. They are found in service organizations such as consultancies, universities and hospitals. Service is not standardized. The time lag between defining a problem and solving it is much shorter than in value chains. You will not find many Google instances because of its specialized nature.

Value Network: focus on the activities that deliver value to customers, rather than the activities of the company (as in VC). Value is non-linearly produced. VN makes no assumptions about primary or secondary, treating intangibles as important as tangibles. Value Networks are found in service organizations such as Banks, travel agencies and IT organizations.

The point about intangibles is critically important. Value Chains have great difficultly expressing intangibles as forms of value. For this reason alone, many would side with the paper mentioned. Carl Stern (as you cite) expressed this brilliantly, prompting many strategists to abandon Value Chains for services once and for all. Porter’s separation of primary and secondary activities implies repetitive and linear behaviour. The nature of services, however, is characterized by heterogeneity and non-linearity, further dissuading service strategists from Value Chains.

Is the answer Value Networks? Only time will tell. Value Network analysis is gaining ground in services modeling. Personally, it feels like a prelude to Systems Thinking/System Dynamics. The ability to perform simulations is a powerful force, moving organizations away from the subjective ontology mentioned.

One last point. The use of the word “chaotic” and “non-linear” seemed inconsistent.

Non-Linear: A process is said to be linear if the process response is proportional to the stimulus given to it. For example, if you double the amount deposited in a conventional savings account then you will receive double the interest. Similarly, if you work ten percent longer hours, you would hope to accomplish ten percent more work. These are linear responses.

Models that assume a process is linear have been extensively studied because the math for such models is relatively straightforward, and linear models can adequately represent the behavior of many realistic processes over a useful range of conditions. It is often possible to solve the equations for linear models without the need to use computers. Thus, the ease of solution for linear models led to their use even in situation where the real-world process was known to be nonlinear. Many business processes are nonlinear, especially when pressed to extremes.

For example, while it may be true that if you work ten percent longer hours you will accomplish ten percent more work, it is probably not true that if you work twice as many hours you will accomplish twice as much work. Many of us have attempted to do this, and have soon suffered from burnout leading to a reduction in our working effectiveness. This is a nonlinear response. Similarly, the available production capacity may limit the amount of a product that can be sold, regardless of the amount of sales effort or the degree of customer demand.

Chaotic: implies randomness or non-determinism. Non-linear (complex) systems are very deterministic. They are characterized by organized complexity rather than disorganized complexity (random systems). IT organizations are complex, not random. This is what Kimberly meant when she pointed out the CMDB doesn’t really help with dynamic complexity. Better and faster data may speed up systems, but on its own doesn’t reveal non-linear behaviours. For this you need to model the larger system. Thus the great interest in Value Networks and System Dynamics.

[alphasong]

Glad to see the debate continue, but I am not necessarily swayed. The more I read about value networks (e.g. Stern), the more I am struck by the starry-eyed dot-bomb tone characteristic of the late 1990s when the concept was being developed. Surely, we might reasonably question some of the theories emergent during that period?

There is no question that at an abstract level, the primary IT value chain is repeatable. I have witnessed its execution through the demand/build/run cycle hundreds of times, resulting in portfolios of thousands of services. After a while, it's just not that complicated. Are there chaotic dynamics? You betcha, and there are processes like Issue Management (for projects) and Incident Management (for operations) to cope with such eventualities.

One common dynamic for which I seek an explanatory paradigm is the continual frustration of IT architects, auditors, and security personnel in achieving goals that are either longer term, or orthogonal to the perspective of the primary demand/build/run cycle. "Those bureaucrats, who never do any Real Work, are trying to tell me how to build this system!" versus "Here we try yet again to redirect another project at risk of building something redundant or ill-conceived." I really can't think of a clearer example of primary versus supporting dynamics.

(For further evidence, suggest you research the debates around Agile Methods.)

If value network were a more appropriate explanatory paradigm, I would expect to NOT see such consistent issues. I would expect the issues to be more heterogenous. But there they are, the same issues, in large IT shop after large IT shop.

I am confused how the concept of tangibles and intangibles is brought in here. I see no reason why a simple value chain analysis is limited to tangibles. Clearly, the utility and value of IT services needs to be measured using both (e.g. a balanced scorecard approach). There is an implication here that the supporting processes are intangible and the primary processes are tangible. Can you show me where Porter states this? Bear in mind, he is my primary source.

You say, "The nature of services, however, is characterized by heterogeneity and non-linearity..." I see the heterogeneity as manifesting at the *instance* level. Generic dynamics can be *abstracted.* (The process concept of Incident Management is nothing but this.) And I continue to detect a certain consistent failure of such abstraction in all of these comments. The discussion seems mired at the instance level. To quote Kimberly, perhaps I'm being a bit unfair. Abstraction is hard to get, perhaps even more so than non-linearity :-)

Finally, re: chaos and non-linearity. Recommend the Wikipedia article on chaos theory. Quote: "In mathematics and physics, chaos theory describes the behavior of certain nonlinear dynamical systems that under certain conditions exhibit a phenomenon known as chaos."

Using chaotic and non-linear interchangeably might be a bit imprecise, but I think it's clear they are closely related. It's those non-linear functions in the aggregate that cause behavior so unpredictable as to *appear* random. (See also the Wikipedia article on Randomness.)

So, to square the circle: I propose that what we have is a repeatable IT Value Chain that manages non-linear and chaotic Value Networks.

[stratman]

Charlie -

You are speaking authoritatively on topics where you are clearly not well-versed. Google and Wikipedia are not substitutes for the real thing.

1) You are mixing chaos and chaos theory. Chaos is about disorder, while the essence of Chaos Theory is order. The central concept of chaos theory is that while it is impossible to exactly predict the state of a system, it is generally quite possible, even easy, to model the overall behaviour of a complex system.

IT organizations are not random or disordered. They may be unpredictable, in that they cannot be expressed in linear equations, but they are not chaos. If they were, we would apply the tools used for random systems such as probability and statistics. This is, of course, as inappropriate as applying linear models.

But if every problem can be abstracted to a nail...

2) Your reference to Agile methods to support primary and secondary activities is contradictory and self-defeating. The central tenet of Agile is about continually adjusting to the non-linearity of development lifecycles, rather than adhering to tightly coupled plan-driven frameworks. It does not attempt to predict the future, but rather make adjustments to it. This is the essence of non-linear frameworks like the Value Network.

The key point to your example ("...redundant and illconceived...") is based on the pressures of intangibles. Calling them primary or secondary serves no purpose. They are every bit as real as the tangibles. Unfortunately, this cannot be expressed as a category of activity in the Value Chain. They do not exist in this model.

3) Your abstraction logic is also flawed. Since you appear familiar with software methodologies, then you are familiar with Anti-Patterns. Your abstraction argument goes from a difficult problem to a bad solution. The fact that it is an abstraction doesn't justify or make it better. Easier to sell, maybe.

4) On the critique of "starry-eyed dot-bomb tone." As Keynes famously replied to a critic, "When the facts change, I change my opinion. What do you do, Sir?"

We have put both dot-com and dot-bomb behind us and these ideas have held up well. In fact, I'm bewildered as to how fiercely you reject them. Is there something else at work here?

You are certainly free to reject or critique these notions. But please don't come back with Google-ized arguments. They make the debate silly and I'm disinclined to respond.

[alphasong]

I'm disinclined to accept your opinions as any better founded than mine. I've not noticed any substantive references to IT governance, IT service management, IT portfolio management, or other significant bodies of theory relevant to this discussion.

Re 1), chaos versus chaos theory is nitpicking. If it makes you feel better, I admit that I have used the term "chaos" where I meant "chaos in the sense of chaos theory," consistently through the thread. If I had meant "chaos in the sense of entropic randomness" I would have used the term "randomness." Nor is my familiarity with chaos theory and nonlinear dynamics restricted to Wikipedia; I simply chose that article as a reasonable common point of reference that others could quickly access.

I agree completely that IT service dynamics may *appear* random but are in fact *not.*

Re: 2) Agile methods. While the dynamics of many (not all) software projects may well be resistant to predictive control (the essential problem Agile is attempting to solve), there is no question that projects are

1) assessed for feasibility and ROI,
2) authorized and funded and executed and
3) expected to turn over a finished product for operational benefit.

When does that value chain sequence not exist in enterprise IT?

I would enjoy seeing you try to convince a project manager (or an Agile developer) of your interesting speculation that governance objectives, as stated by parties external to the core project team, are "every bit as real as the tangibles." I've been in that conversation multitudes of times in my career. It is invariably a difficult one.

And I fail to understand the claim that "they [governance objectives] do not exist in [the Value Chain] model." I am talking about my IT value chain, not Michael Porter's generic manufacturing one. In my IT value chain, IT governance is clearly represented as a supporting activity. (By the way, I am not the only one working with the IT value chain concept. I have comparable examples from the Brazilian consultancy Sincrono, from the U.K. ITIL theorist John Gibert, and from Mark Lutchen, a senior PWC partner and former CIO. Paul Strassmann also has written about the IT value chain.)

Your comments in 3) boil down to nothing more than your asserting that I am wrong. More substance please. In what particular manner is the logic flawed? In what ways are the model's explanatory powers sub-optimal?

Finally 4) I am not fiercely rejecting the concept of value network. I have proposed that we have a value chain that manages value networks. Where I draw the line is in accepting that value networks have inherently superior explanatory powers for all aspects of enterprise IT. On the erp4it list, Martin Erb has also chimed in supporting this point of view. What is your response?