Despite having 3,000 more words to write about Arthur Duke of Wellington’s influence on contemporary intelligence operations, I had some thoughts regarding a blog post by my friend David Eaves. You can read it here.
In it he describes the Ontario Government’s difficulty in fostering an organizational system deemed a “network.” David posits that the primary sticking point is that networks succeed by virtue of free-agency, wherein an individual nodes connections are increased based on reciprocal inclusion in the network. If a node is instructed to join the network its connections are confined by this mandate.
The interesting thing I observe in both the remarks by David, Dana Richardson and many other corporate leaders is that they say they have difficulty in fostering a network, but they (from what I’ve seen) don’t have any serious metrics for success.
What is the merit of a network? Networks are useful because they are comparatively “strong.” In fact, the measure of strength in any network is the proportion of node failure that leads to systemic breakdown of the network. Organizations under threat benefit most from this type of structure. More institutional organizations significantly less so.
Even basic business process modelling will demonstrate that most businesses take a huge hit to efficiency when configured in an enmeshed lateral system that most people view as a network. Mechanistic bureaucracy works really well - rationalization has a much firmer footing in efficiency than does networking.
Networking is, in most people’s imaginings, a way of fostering innovation. By bringing people from different kinds of expertise and backgrounds they hope it will foster cross-cutting ideas that will result in new and interesting cooperative innovations that will increase overall efficiency. Afterall, most innovation is the combination of two distinct things into a single new thing, or the separation of a single thing into two different things.
Without postulating too much about the overselling of innovation in corporate culture let’s go back to the seeming “lack of success” in governmental networks. It seems that it would benefit much more from service oriented architecture (SOA) perspective than an emergent network. SOA, for all those not in the jargon-drenched world of enterprise business modelling, is a method of envisioning all activities of an organization as business processes packaged as services. A compositional breakdown of each service reveals agents and exchanges or messages.
Agents represent the people and external machines of a service, exchanges avenues of interchange of information, money, etc, messages are like exchanges except there is no interchange - objects flow in only one direction.
While networks are involved the network emerges as a result of the various services and business transaction available as a result. This is the logical conclusion to “core competency” based strategy. SOA is ripped straight from computer engineering textbooks that discuss the idea of object-oriented programming, the idea that a computer system can be broken down into encapsulated objects that manipulate data based on programmed behaviour.
This is opposed to languages which script every interaction procedure, so called procedural languages. Bureaucracies are best described in terms of procedural programming, whereas what most people envision as networks are best articulated using object-oriented languages.
An efficient network in an organization like a governmental body emerges when each node makes explicit the information it retains and services (with associated contracts) it can offer. A service contract stipulates the input required to successfully transact a given service including information, funding, external services etc. When these attributes are readily discoverable and activated by other nodes you have SOA - which isn’t a strictly emergent network per se, since the attributes of a given node are assigned.
In more concrete terms this makes it easy to find the “go-to” person on a given subject, establish reliable trust networks and detect organizational clusters that can be optimized. Those transaction pathways most commonly traveled should be made more efficient (roads most traveled should become freeways). It also makes pruning simpler since a node’s value is determined by a ratio of incoming traffic over cost of activity.
By simply making government employees generate facebook pages, it wastes my tax money. The tools must reflect the structure one is seeking in an architected environment. Making everyone learn and use Facebook is akin to making everyone learn and use a Nintendo DS.
Having a clear idea of what benefits one seeks from a network is the only way you can see success and judiciously picking tools that propel those benefits is key. When you fulfill those requirements and communicate your findings to the organization - the network will grow without hierarchal impetus.