The Sources of Innovation (Framework)
The Sources of Innovation framework provides the analytical architecture for predicting which actor in an industry's value chain will originate its innovations. The framework identifies three principal sources — users, manufacturers, and suppliers — and specifies the structural conditions under which each source dominates. The conditions are economic rather than cultural: the source of innovation in any industry is the actor who can capture the greatest share of the innovation's benefit relative to its cost. When users capture the benefit, users innovate. When manufacturers capture it, manufacturers innovate. When suppliers capture it, suppliers innovate.
In the AI Story
The framework emerged from von Hippel's effort to generalize the empirical findings from scientific instruments into a predictive theory. The 1988 book The Sources of Innovation presented the full framework with case studies demonstrating how it applied across diverse industries. The core prediction — that innovation sources are determined by the distribution of innovation benefits — made user innovation legible as a structural rather than accidental phenomenon.
The framework has predictive power because the distribution of innovation benefits can be analyzed before innovations occur. Industries where users capture large benefits from customization — scientific research, surgical practice, elite sporting performance — are structurally favorable to user innovation. Industries where manufacturers capture large benefits from standardization — consumer packaged goods, commodity chemicals, mass-market electronics — are structurally favorable to manufacturer innovation. Suppliers dominate innovation where they capture benefits through process improvements that manufacturers cannot easily replicate.
The AI moment does not change the underlying logic of the framework but shifts the parameters dramatically. When user innovation cost drops by orders of magnitude, the benefit-cost ratio for user innovation improves correspondingly. Industries that were marginally in manufacturer territory move into user territory. Industries that were firmly in user territory become even more dominated by users. The framework predicts the direction of shift with considerable precision.
Origin
The Sources of Innovation (Oxford University Press, 1988) synthesized more than a decade of empirical research into a unified framework. The book's methodological contribution was to demonstrate that innovation source was a dependent variable — something to be explained — rather than an assumption to be made. Its theoretical contribution was the benefit-capture model that explained why different sources dominated different industries.
The framework established the research program that produced Democratizing Innovation (2005) and Free Innovation (2017). Each subsequent book extended the framework into new domains and new questions, but the 1988 architecture remained the analytical foundation.
Key Ideas
Three principal sources. Users, manufacturers, and suppliers each dominate innovation in different industries based on structural conditions.
Benefit-capture determines source. The actor who can capture the greatest share of innovation benefit relative to cost is the actor who innovates.
Predictable rather than accidental. The framework transforms innovation source from an assumption into an analyzable variable with identifiable conditions.
Industries classifiable. Scientific research and elite sport favor user innovation; consumer packaged goods favor manufacturer innovation; process-intensive industries often favor supplier innovation.
AI shifts parameters. The language interface alters the benefit-cost ratio for user innovation, shifting industries toward user-dominated patterns.
Appears in the Orange Pill Cycle
Further reading
- Eric von Hippel, The Sources of Innovation (Oxford University Press, 1988, open access)
- Joel Mokyr, The Lever of Riches (Oxford University Press, 1990)
- Wesley Cohen and Daniel Levinthal, 'Absorptive Capacity' (Administrative Science Quarterly, 1990)
- Eric von Hippel, 'The Impact of Sticky Local Information on the Locus of Innovation' (Management Science, 1998)