It is a simple question really. What do you have to do yourself – and what can you outsources? Set aside for now the strategic interests of owning the results and the bandwidth issues of whether or not you have the time, these will be topics for a different discussion. The question arrises in every size firm, from a single person, a small team or an entire organization.
Let me propose a simple two dimensional approach to tease out the factors weighing upon the degree to which innovation can be outsourced. The axes are simple. The independent axis is the degree to which a problem can be constrained or stated clearly. Low constrained questions are very open and nebulous, while highly constrained questions are detailed with little room for any creativity. On the vertical axis we have “capital and capabilities”. These factors are lumped together to generalize the discussion, but should be taken broadly to mean education, equipment and resources considered to be human or other forms of capital equipment. Low capital/capabilities would be a well trained monkey, and high capital would be something like a team of geniuses with billion dollar budgets. Some examples will quickly make this clear.
In the high/high quadrant we find large public works projects like roads and bridge building. Huge amounts of capital are required, both in design expertise in engineering for things like bridges, and also the equipment and budget required to build them. Ultimately the constraints are very specific though.
In the High capital, low constraint we see lots of the planning for long term government projects. Here we find things like “moonshot” where the US was determined to get a man to the moon and back, but it was not clear how. This quadrant is the domain of NASA, NIH, CERN, cancer research, the human genome project and so forth. Many projects start here, and may move to the High/high quadrant when they are ready for implementation.
In the low capital, high constraint. This is perhaps the most interesting quadrant. This is the realm of incentive where single people in distributed fashion are solving problems if given enough direction. Forces enabling smaller and smaller firmst doing more distributed tasks include the increasing capabilities at low capital expenditure. For example, the computer, communications, advanced software and access to information make a “firm of one” sitting in a windowless room with only a computer terminal and a broadband conenction, very productive.
Low capital, low constraint is a funny quadrant, because it involves untrained participants solving poorly defined problems. I call this the failure detection quadrant.
What keeps things un-outsourceable are a few forces. Unintended consequences, or highly interdependent components, and disparate required disciplinary expertise all reduce the outsourceability of a problem.
In our increasingly interconnected society, and interdisciplinary technology landscape – how will firms be structured in the modern economy? With such a broad question it may be tough to come to a singular answer, but these forces help to create a framework for analysis.