Innovation Networks: a la Google or a la Apple?

A recent article in International Herald Tribune entitled "Yin and Yang of corporate creativity" describes two approaches to innovation, one of Apple and another of Google. The Google approach is a bottom up, open innovation which is based on rapid experimentation and receiving quick customer feedback. The Apple model is more top down model where the company achieves close to perfect integration of different elements in the product and then pushes it down to customers (remember a now famous comment from Jobs: it's not the consumers' job to know what they want).

This dichotomy is very much relevant to the issues of managing innovation networks in an organization or across organizations. Such network can consist of individual collaborators (if this is an inter-personal network) or partner companies (if this is an inter-organizational network). Google's approach is similar to a network where the central member in a network relies on the peripheral members for insights into the environmental dynamics and then works with them to find a solution to the environment's needs and then constantly make small adjustments to the resulting product.

Apple's approach is similar to a network where the central member controls the network and imposes its views on the peripheral members. In such network, innovation occurs in discrete steps with major adjustments embedded in different generation products.

Which network is better for generating innovation? The answer is that it depends on the environment these networks are facing. If success in a given environment depends on coordination of very different elements (e.g. hardware and software) or connecting ideas from very different domains (e.g. music and technology), then top down model of network management a la Apple makes a lot of sense. If success in an environment is based on coordination of similar elements (e.g. only hardware or only software) or ideas from similar domains (e.g. technology only) then the model a la Google makes more sense. Ideally, of course, innovation networks should probably combine two approaches- top down and bottom up- it seems to me that this was the Microsoft's approach in the 1990s.