Systemic vs. Autonomous Innovation
This framework attempts to answer the question:
What is the optimum place for innovation within the business?
The methodology below illustrates how executive management could approach this question.
First, it is paramount that the Company correctly identifies the type of innovation they are dealing with, that is either
Autonomous or Systemic.
In most cases, large corporations operating in high velocity markets (eg. Danaher, Intel, P&G etc) face a situation of autonomous innovation as various business units tend to adapt to new technology advances or societal future trends.
Before deciding where (the optimum place) to effectuate strategic innovation (say digital transformation or a new MVP), the management needs to ascertain what are the levers that make innovation a reality. Based on these levers, the leadership can decide on the locus for innovation, that is:
at the edge of the Company
inside the Company
outside the Company (with appropriate linkages)
And so,
The distinction between autonomous and systemic innovation is paramount to the choice of organizational design.
For instance, when the auto industry changed from drum brakes to disc brakes (autonomous innovation) GM was unable for a while to adopt the new product because the old technology (drum) was already vertically integrated from a production perspective.
The difference between GM and its competitors was the fact that GM was at the time quite centralized while their competitors less so and therefore able to rely on external suppliers and a set of incentives, to “effectuate new combinations “ (Schumpeter).
More generally put:
Autonomous innovation in a Centralized business is a bad combination.
When innovation is systemic in nature, the elements in box 2 play a very important role: the members are dependent on other members (of over they have no control) to deliver (which is underpinned by a culture of collaboration). Hence,
The wrong organizational choice can be costly in innovation.
Once the dissemination between autonomous and systemic innovation takes place, the Company can progress to box 3 and make a better and informed decision.
In other words, the Company has to think in multidimensionally: what is the type of innovation we are engaged in? Is our current organizational structure aligned with the nature of the innovation? And, what is the most optimum place for innovation? Let us consider the following construct (A-D-E):
Type = Autonomous innovation
Organizational design = Decentralized business
Place = Innovate at the Edge
As a rule of thumb, when innovation is autonomous, innovating at the edge could be a scenario suitable for a decentralized business. The major disadvantage of innovating at the edge is coming under fire when the organization engages in operational effectiveness, ie, Company's leadership tends to knock down the “business satellites” first.
Notwithstanding, the combination A-D-E is in most cases, the right choice for a decentralized, (mildly to highly) diversified business.
One example of A-D-E arrangement is Intel’s recent acquisition of Mobileye which provided Intel with the ability to innovate in the driverless car technology arena. In fact, if we were to dig deeper, Intel collaborated with Mobileye well before it acquired the Israeli firm.. ie. Intel started to innovate on driverless car technology "outside the company", box 3C.
When innovation is systemic (ie, it depends on a series of interdependent innovations), decentralized companies are rarely in the position to coordinate properly at the boundaries (fuzzy) and stitch those interdependent innovations together. This is generally speaking more difficult to “effectuate” and poses the danger of value seepage at the boundaries between BU’s or verticals.
Essentially, placing and dealing with innovation, especially in large multi-divisional/multi-vertical and decentralized firms, becomes a question of scale, integration and leadership.
Therefore, companies that are able to integrate autonomous innovation and avoid value seepage (via the right type of org design), can create enormous distinctive advantage.
Acknowledgments
The author would like to thanks Prof. David Collis (Harvard Business School) and Chris Zook (Bain & Co) for their contributions and commentary.
EndFragment
