Loading field map
Innovation is inherently uncertain, non-linear, and difficult to steer. Firms that try to manage it systematically face a persistent tension: they need structure to coordinate effort, yet too much structure can stifle the creativity and flexibility that innovation demands. Over the past seven decades, the subfield of innovation management has produced a series of frameworks that grapple with this tension. Each framework emerged from a specific practical pressure—whether the need to predict adoption, to discipline product development, to explain national differences in technological performance, or to harness external knowledge. The frameworks have not simply replaced one another; many remain in active use, often in hybrid forms, and the field today is marked by productive pluralism.
The earliest frameworks treated innovation as a sequential process. Linear Innovation Models, dominant from the 1950s through the 1980s, depicted innovation as flowing from basic research to applied development to production and finally to market. Two variants coexisted: the technology-push model, in which scientific discovery drives the pipeline, and the market-pull model, in which customer needs trigger development. Both assumed a one-way, orderly progression. This sequential logic was appealing for planning but ignored the feedback loops, setbacks, and social dynamics that real innovation involves.
Diffusion of Innovations, introduced by Everett Rogers in 1962, shifted attention from the firm’s internal pipeline to the social process by which innovations spread across a population. Rogers synthesized hundreds of studies to show that adoption follows an S-curve, driven by communication channels, time, and the social system. The framework introduced categories of adopters (innovators, early majority, late majority, laggards) and highlighted the role of opinion leaders. Where linear models focused on producing the innovation, diffusion theory explained why some innovations succeed in the marketplace while others languish. The two frameworks coexisted for decades: firms used linear models to organize development and diffusion theory to plan market entry. Diffusion’s emphasis on network effects and social contagion later provided a foundation for thinking about innovation ecosystems.
By the 1980s, managers wanted more control over the messy front end of innovation. Stage-Gate Process, introduced by Robert Cooper in 1986, absorbed the linear logic of the pipeline but added decision points—gates—where projects could be killed or redirected. Each stage (idea screening, business case, development, testing, launch) ends with a gate where criteria are applied. Stage-Gate did not reject the linear model; it disciplined it. The framework became the dominant product-development template in large firms, though critics argued it could slow down radical innovation. Today, Stage-Gate is often blended with agile methods in hybrid “Agile-Stage-Gate” models.
At roughly the same time, National Innovation Systems (NIS), articulated by Christopher Freeman and Bengt-Åke Lundvall in the late 1980s, moved the unit of analysis from the firm to the nation. NIS argued that a country’s innovative performance depends on the institutional infrastructure: universities, research labs, patent laws, funding agencies, and the interactions among them. This framework contrasted sharply with the firm-centric linear models and Stage-Gate. It showed that innovation is not just a managerial process but a systemic outcome shaped by policy and history. NIS later extended into regional and sectoral innovation systems, and its institutional focus prepared the ground for the more strategic, relational view of innovation ecosystems.
The mid-1990s brought three frameworks that addressed why firms with good processes still struggle. Knowledge-Based View (KBV), emerging from the resource-based view of strategy, treated knowledge as the most strategically important resource. Firms innovate by creating, transferring, and integrating knowledge. KBV provided a bridge from strategy to innovation management: it explained why some firms are better at innovation (they have superior knowledge assets and routines) and why collaboration is difficult (knowledge is sticky and tacit). This framework later became a foundation for Open Innovation, which explicitly contrasts internal knowledge generation with external sourcing.
Disruptive Innovation, coined by Clayton Christensen in 1997, offered a different explanation of failure. Incumbents with good capabilities and attentive customers can still be overthrown by entrants that target overlooked segments with simpler, cheaper offerings and then move upmarket. Disruption is not about poor management but about the logic of resource allocation and customer focus. This framework stood in tension with Dynamic Capabilities: good capabilities may not prevent disruption if they are aligned with the wrong market trajectory.
Dynamic Capabilities, developed by David Teece and others in the late 1990s, focused on how firms sense new opportunities, seize them, and transform their resources. Unlike the static resource-based view, dynamic capabilities emphasized change and adaptation. The framework absorbed ideas from organizational learning and evolutionary economics. It provided a vocabulary for why some firms repeatedly innovate: they have routines for reconfiguring assets. Dynamic Capabilities and Disruptive Innovation remain in live disagreement: one stresses the firm’s ability to adapt, the other stresses structural forces that overwhelm adaptation. Both frameworks, however, shifted attention from managing single projects to managing the firm’s long-term innovative capacity.
The turn of the millennium saw a wave of frameworks that challenged the closed, firm-centric assumptions of earlier models. Innovation Ecosystems, popularized by Ron Adner and others from 2000 onward, argued that innovation success depends on the alignment of multiple interdependent actors—suppliers, complements, regulators, customers. The unit of analysis is the ecosystem, not the firm or the nation. This framework extended the network logic of Diffusion of Innovations and the institutional logic of National Innovation Systems into a strategic-relational view. Where NIS emphasized government policy, ecosystems emphasize the firm’s ability to orchestrate partners. Ecosystems also absorbed the interdependence that Stage-Gate and linear models had ignored.
Agile Innovation, emerging from software development practices after the 2001 Agile Manifesto, introduced iterative, cross-functional, customer-responsive development. Agile is more a set of practices than a formal theory, but it qualifies as a framework because it embodies a distinctive logic: break work into short cycles, test assumptions early, and adapt. Agile challenged the sequential, plan-driven logic of Stage-Gate. Yet the two have been combined in practice: firms use Stage-Gate for portfolio governance and Agile for execution. Agile’s emphasis on rapid sensing and adapting aligns closely with Dynamic Capabilities, and many scholars now see Agile as a practical instantiation of those capabilities.
Open Innovation, formalized by Henry Chesbrough in 2003, directly challenged the closed innovation model in which firms generate and commercialize ideas internally. Open Innovation argues that firms should use both internal and external ideas and paths to market. It absorbed the Knowledge-Based View’s insight that valuable knowledge is widely distributed, and it extended Diffusion of Innovations by treating external adoption as a deliberate strategy. Open Innovation has become one of the most influential frameworks in the field, with firms adopting practices such as inbound licensing, spin-offs, and innovation intermediaries. It coexists with Innovation Ecosystems: ecosystems describe the structure of interdependence, while Open Innovation describes the strategy of boundary-spanning.
Today, no single framework dominates innovation management. The leading frameworks—Open Innovation, Innovation Ecosystems, Agile Innovation, Dynamic Capabilities, and Diffusion of Innovations—are all actively used, often in combination. They agree on several points: innovation is inherently uncertain and requires iterative learning; external knowledge and partnerships are critical; and firms must balance structure with flexibility. They disagree on the primary unit of analysis (firm, ecosystem, or system), on whether disruption is a predictable threat or a manageable risk, and on the degree to which innovation can be systematized without losing its creative edge.
Earlier frameworks have not disappeared. Linear models survive in simplified project planning. Stage-Gate remains the backbone of many product-development processes, now hybridized with Agile. National Innovation Systems informs science and technology policy. The Knowledge-Based View underpins research on absorptive capacity and knowledge transfer. Disruptive Innovation continues to shape strategy teaching and startup narratives. Dynamic Capabilities provides a meta-framework for understanding organizational adaptation. The field’s history is not a story of replacement but of accumulation and recombination, with each framework adding a lens that later practitioners and scholars can draw on as the context demands.