In a perfect world, the four-step translation of strategic direction into military capability model should deliver a force structure able to implement extant defence policy. However, the analysis laid out in this book shows that this does not always happen. Not unexpectedly, there is no single reason why. The model is affected by untold factors, some of which are initiated by the government or defence decision makers themselves, others are determined by outside agencies. These include: politicians; elements of MoD head office, the single services and delivery agents; the defence industry; the media; the public; allies; and potential state and non-state adversaries. Except for the inputs of potential adversaries, most of the factors that these agencies introduce are likely to be in pursuit of valid and reasonable effects; nevertheless, they all add complexity to the model and make the task of providing fit-for-purpose armed forces considerably more challenging.

We know the translation of strategic direction into military capability is an open-ended activity. The introduction of uncontrollable and disruptive factors also make it complex and intractable. It has all the hallmarks of a wicked problem (see Rittel and Webber, 1973; and Head, 2008). No single action will resolve all the adverse effects of the events and influences that impact strategic direction into military capability activity; moreover, any attempt to mitigate the complexity that these factors introduce must be mindful of the following two points. In the first instance, an improvement in one aspect of the activity may well lead to a regression elsewhere. For example, several senior officials believed the freedoms within which the delegated model were implemented allowed the single services to introduce customized processes that best tackled the specific capability management demands within their environment. However, that resulted in considerable inefficiency within MoD head office and the delivery agents who then had to align with four bespoke operating models.