by Philip Boxer
Charlie, Thank you for your thoughts on value-driven architecture relating to the Governance paper, and for drawing my attention to your paper on linking product strategy with architecture. Your use of “context-specific challenges” nicely identified the nature of the challenge presented by asymmetric forms of demand. A close look at your paper identified a number of other points of connection – particularly since you start from Michael Porter’s Competitive Advantage (M.E. Porter, Free Press, 1985). Porter’s concepts have been a jumping off point for me too, although I think there are a number of difficulties in using his concepts in relation to our third (demand) asymmetry.
I have always found his use of ‘horizontal’ and ‘vertical’ confusing (horizontal linkages = vertical integration = integration along the whole length of the value chain viz aluminum foil). The use of vertical linkages between entities in different systems corresponds much more to Christensen’s concept of ‘value network’ (The Innovator’s Challenge, Harvard Business School Press, 1997), i.e. the particular way in which technologies organize the design of products and services, and which I refer to as an ‘effects ladder’ (viz the way we all used to agree how to use disc drives).
In a stable industry, because the variations in value context can be ignored (there used not to be much variation in the way disc drives were used), the effects ladder defines the organization of the value cluster, describing its business model. But as the levels of variation increase, so that the value models within the value cluster become increasingly heterogeneous, so too does the variation increase in the organization of the effects ladders within the value cluster. Our question is whether it is useful to look at the demand-side variation independently of the supply-side.
What is interesting about your paper is that the organization of a value cluster is a response to the value drivers – “a value model captures the drivers that determine how satisfied a particular market segment is, and how difficult it will be to satisfy them.” At the level of the particular demands facing a value model, I can see how this enables you to concentrate on the particular architecture challenges presented (viz designing a car that stops under differing conditions of use).
In contrast, we have placed much greater emphasis on defining the effects ladder for the customer – i.e. the customer’s value model – independently of the value model of the supplier (viz how can customers define how they want to use ICT within their business independently of the particular technology solutions they might subsequently use). Hence our approach of separating out the demand-side risks associated with embedding services in the customer’s value model, and considering the supply-side risks of delivering particular services out of an infrastructure separately. This is because we have wanted to see to what extent the demand-side value drivers (idiosyncratic ways in which users define their needs) diverge from supply-side value drivers (limitations in the way the technology can be used).
Of course, both approaches are needed, depending on how tightly coupled are the architectures of the supply- and demand-sides. But the whole separating out of the transactional and transformational elements of the value model from its tacit elements (The next revolution in transactions, McKinsey Quarterly No 4, 2005) suggests that this uncoupling process is increasing. What is your sense of this?
next in this thread
by Philip Boxer