Theory-Driven Innovation in Organizations: From Combinatorial Possibilities to Practical Breakthroughs

Listen to this article:

Abstract: Organizations face astronomical numbers of potential innovation pathways, yet most successfully navigate toward useful combinations of ideas, technologies, and processes. This article examines how theory-driven experimentation generates combinatorial salience within organizational contexts, enabling practitioners to identify promising innovations among indefinite possibilities. Drawing on recent advances in combinatorial innovation theory and cognitive science, we argue that organizational innovation depends on the capacity of organizational actors to theorize, reason causally, and experiment systematically. Through examination of contemporary organizational cases spanning healthcare, manufacturing, and technology sectors, we identify evidence-based interventions for building theory-driven innovation capacity. The article contributes to practice by offering actionable strategies for cultivating organizational environments where theory-laden experimentation accelerates learning cycles and enables discovery of novel yet feasible innovations.

Innovation leaders face a fundamental paradox. The digital age offers unprecedented access to knowledge and technologies—yet the sheer magnitude of possibilities often paralyzes rather than propels action. A modest innovation initiative combining just 50 existing organizational capabilities yields over a quadrillion possible configurations. The practical question becomes urgent: How do organizations successfully navigate this combinatorial explosion to arrive at useful innovations?

Traditional approaches emphasize recombination—mixing and matching existing components into novel configurations (Arthur, 2009; Weitzman, 1998). While combinatorial theory explains long-run technological evolution, it leaves critical questions unanswered for practitioners: Which combinations should teams pursue first? How do certain possibilities become salient while countless others remain invisible?

Recent theoretical advances offer fresh insight. Felin and Singell (2025) argue that theory-driven experimentation generates combinatorial salience by providing shortcuts for brute force search—making the otherwise intractable innovation landscape analytically manageable. Rather than portraying organizational innovation as blind tinkering, this perspective recognizes that human actors theorize, reason causally, and design experiments that deliberately test hypotheses about what might work and why.

For organizations, these insights carry profound implications. If innovation depends fundamentally on employees' capacity to theorize and experiment systematically, then building innovation capability means cultivating environments where such reasoning flourishes.

The Organizational Innovation Landscape

Defining Theory-Driven Innovation in Practice

Theory-driven innovation refers to organizational processes where actors deliberately formulate hypotheses about cause-effect relationships, design experiments to test those hypotheses, and use evidence to refine both their theories and their actions (Camuffo et al., 2020). This contrasts with common portrayals of innovation as serendipitous tinkering or pure trial-and-error.

In practical terms, theory-driven innovation involves several interconnected capabilities. Problem formulation requires clearly articulating the gap between current and desired states. Hypothesis generation involves developing testable propositions about which combinations of elements will produce desired outcomes. Systematic experimentation entails designing rigorous tests that can discriminate between competing explanations.

As Felin and Singell (2025) emphasize, just as the Wright brothers systematically reasoned through problems of lift, propulsion, and steering rather than randomly combining components, contemporary organizational innovators identify relevant combinations through deliberate causal reasoning rather than exhaustive search.

State of Practice

Contemporary organizational innovation practices reveal considerable heterogeneity. Some firms embrace highly structured, scientifically-informed approaches. Procter & Gamble's "Connect + Develop" program employs systematic frameworks for identifying external innovations addressing clearly specified internal needs (Huston & Sakkab, 2006). IDEO's design thinking methodology incorporates theory-driven experimentation through rapid prototyping cycles testing specific assumptions (Brown, 2008).

At the other extreme, many organizations operate under the "fail fast" paradigm—encouraging extensive experimentation with minimal theoretical grounding, potentially wasting resources testing combinations that rudimentary causal reasoning would have eliminated.

Recent empirical evidence suggests theory-driven approaches can deliver substantial performance advantages. Camuffo et al.'s (2020) randomized controlled trial with 116 Italian startups demonstrated that firms trained in scientific decision-making achieved revenue growth rates significantly higher than control groups and were more likely to pivot successfully when initial theories proved incorrect. Yet adoption remains limited, often reflecting organizational architectures that inhibit rather than enable theory-driven experimentation.

Organizational and Individual Consequences

Performance Impacts

The choice between theory-driven and undirected innovation approaches generates measurable organizational consequences. Development cycle time represents the most immediately visible impact—organizations employing systematic hypothesis testing may reduce time-to-market through improved search efficiency (Eisenhardt & Tabrizi, 1995).

Resource utilization efficiency shows similarly important effects. Research suggests that teams trained in scientific decision-making can achieve higher success rates per dollar invested, primarily by avoiding experiments that theoretical analysis suggests would fail (Garnier, 2008).

Innovation quality metrics may further differentiate approaches. Patent citation analysis reveals that innovations emerging from more systematic processes can receive higher forward citations, suggesting potentially greater technological impact (Fleming & Sorenson, 2004).

When Moderna developed its COVID-19 vaccine in just 42 days, the speed reflected years of prior theory-building about mRNA mechanisms and systematic experimentation establishing causal relationships—illustrating how accumulated theoretical knowledge can accelerate response to novel challenges.

Individual and Team Impacts

Beyond organizational metrics, innovation approaches profoundly affect individual employees. Teams explicitly trained in hypothesis-driven experimentation may report higher psychological safety scores, as organizations emphasizing theory testing reduce perceived risk in acknowledging uncertainty (Edmondson, 1999).

Pixar Animation Studios exemplifies this dynamic. Their "Braintrust" process explicitly separates hypothesis evaluation from personal judgment, encouraging directors to present theories and inviting constructive challenge based on evidence rather than hierarchy (Catmull, 2014).

Employees practicing theory-driven experimentation may develop transferable problem-solving capabilities applicable across diverse challenges (Camuffo et al., 2020). Amazon's decision framework illustrates how structured hypothesis-testing can enhance employee autonomy—reversible experiments receive minimal oversight provided teams clearly articulate hypotheses and success criteria (Bryar & Carr, 2021).

Evidence-Based Organizational Responses

Structured Hypothesis Development Frameworks

Organizations can improve innovation outcomes by implementing systematic frameworks guiding employees through theory formation. Research demonstrates that structured hypothesis development can help teams identify viable innovation pathways faster and with fewer failed experiments (Camuffo et al., 2020).

Effective frameworks incorporate several core elements. Problem decomposition requires breaking complex challenges into tractable sub-problems. Explicit assumption surfacing makes tacit beliefs testable. Hypothesis specification protocols guide teams from general theories to testable predictions with clear success criteria.

Intuit demonstrates these principles. The company trained employees in hypothesis-driven experimentation, requiring teams to articulate specific predictions before launching initiatives. A mobile payment project initially hypothesized that small merchants would value improved transaction speed. Early experiments revealed unexpected value in simplified reconciliation instead, enabling evidence-based pivoting toward a successful merchant services platform.

Approaches include hypothesis canvas tools, theory-building workshops, hypothesis review processes, and assumption testing sprints focused on validating critical beliefs before major resource commitments.

Systematic Experimentation Architecture

While hypothesis formation provides direction, organizations must build capabilities for rigorous testing. Evidence suggests systematic experimentation architectures can improve both learning speed and innovation success rates (Thomke & Manzi, 2014).

Effective architectures incorporate clear learning objectives, appropriate rigor levels matched to experimental stakes, and rapid iteration capability enabling learning through successive refinement.

Amazon exemplifies systematic experimentation at scale, running thousands of controlled experiments annually testing hypotheses about user interfaces, algorithms, pricing, and operations. Their platform provides standardized tools for hypothesis specification, randomization, and analysis (Kohavi et al., 2020). Critically, Amazon values failed experiments that disconfirm hypotheses equally with successful ones, provided they generate clear learning.

Boston Scientific applied similar principles in medical device development, requiring research teams to articulate causal theories before initiating projects, with planned experiments testing critical assumptions at each development stage.

Approaches include experimentation platforms, experiment review protocols, rapid prototyping capabilities, data infrastructure, and statistical literacy programs.

Psychological Safety and Challenge Norms

Theory-driven innovation depends critically on employees' willingness to propose uncertain hypotheses and challenge prevailing theories. Meta-analytic evidence suggests psychological safety correlates with innovation performance (Frazier et al., 2017).

Building psychological safety requires deliberate leadership action. Modeling hypothesis uncertainty—when leaders explicitly frame their thinking as hypothesis rather than certainty—signals that uncertainty is valued. Rewarding thoughtful experimentation distinguishes between poorly-designed experiments and well-designed experiments that disconfirm hypotheses.

Microsoft undertook deliberate psychological safety building as part of its cultural transformation under Satya Nadella. Leadership introduced "learn-it-all" versus "know-it-all" language, began meetings by sharing uncertainties, and implemented growth mindset training emphasizing learning from failed experiments (Nadella, 2017).

Mayo Clinic addressed psychological safety in clinical innovation through structured discussions where teams articulate treatment theories, identify uncertainties, and design low-risk tests—with protocols protecting clinicians who report unexpected outcomes from professional liability concerns.

Approaches include leadership communication training, after-action review protocols, anonymous challenge channels, failure analysis frameworks, and psychological safety assessments.

Distributed Problem-Solving and Theoretical Diversity

Organizations can enhance theory-driven innovation by cultivating diverse theoretical perspectives. While demographic diversity shows inconsistent effects, cognitive diversity—variation in how individuals frame problems and generate hypotheses—can predict creative outcomes (Cronin & Weingart, 2007). Teams with diverse theoretical perspectives may identify more viable solution pathways (Hong & Page, 2004).

IDEO systematically leverages theoretical diversity through multidisciplinary teams deliberately including designers, engineers, anthropologists, and business strategists (Brown, 2008). On a healthcare project addressing medication adherence, different disciplines theorized that technology reminders, social support, or economic incentives would dominate—leading teams to test each hypothesis and discover that all three mechanisms mattered in different patient segments.

Procter & Gamble institutionalized external perspective integration through Connect + Develop, recognizing that internal teams developed theoretically narrow hypotheses. External solutions often embodied theories internal teams hadn't considered (Huston & Sakkab, 2006).

Approaches include cognitive diversity mapping, perspective-taking protocols, external hypothesis sourcing, competing hypothesis frameworks, and theoretical integration workshops.

Capability Building and Scientific Reasoning

Individual-level capabilities ultimately determine whether theory-driven innovation succeeds. Evidence demonstrates that scientific reasoning capabilities are trainable and can yield measurable performance improvements (Camuffo et al., 2020).

Effective programs incorporate action-oriented learning through practice on real challenges, scaffolded progression allowing skill development over time, and distributed expertise recognizing complementary capabilities across teams.

General Electric pursued capability development through its "FastWorks" initiative, training thousands of employees in hypothesis-driven development (Ries, 2011). For a new ultrasound device, teams articulated specific hypotheses about valued features, developed minimum viable products testing those hypotheses, and iterated based on evidence.

Approaches include embedded learning programs, hypothesis mentorship, experimental design workshops, evidence interpretation training, and communities of practice.

Building Long-Term Innovation Capability

Organizational Learning Systems

Theory-driven innovation generates value through accumulation of reusable knowledge over time. Companies with more formal knowledge management practices may achieve higher innovation productivity (Smith et al., 2005).

Effective systems require theory-level capture documenting underlying causal reasoning rather than merely outcomes, structured codification making tacit knowledge explicit, and active knowledge transfer through brokers and forums.

Toyota pioneered organizational learning through its A3 problem-solving approach, with one-page documents capturing problem definition, root cause hypothesis, experiments, results, and theoretical insights (Liker, 2004). Over decades, Toyota accumulated thousands of A3s representing organizational knowledge about manufacturing causality.

Approaches include hypothesis libraries, causal mechanism databases, experiment review protocols, knowledge broker roles, and learning-oriented project launches.

Leadership Mindsets and Role Modeling

Sustained theory-driven innovation depends on leadership mindsets and behaviors. Effective leadership requires embracing uncertainty, valuing learning over advocacy, and distinguishing process from outcome quality.

Satya Nadella's leadership at Microsoft exemplifies these principles. Upon becoming CEO, Nadella confronted a culture expecting leaders to have answers. Through consistent role modeling—framing strategic moves as hypotheses, sharing failed experiments, revising theories based on evidence—he enabled cultural evolution toward experimental approaches (Nadella, 2017).

Approaches include leadership communication development, hypothesis review participation, public theory revision, learning-oriented performance dialogue, and visible experimentation by leaders.

Adaptive Governance

Sustained theory-driven innovation requires institutional architecture supporting systematic experimentation. Research highlights potential misalignments between traditional governance and experimental approaches (Cooper, 2008).

Adaptive governance incorporates hypothesis-based resource allocation funding experiments rather than predetermined solutions, progressive commitment tying funding to validated learning, and experimental metrics measuring rigor and knowledge accumulation.

Amazon's governance system distinguishes between consequential decisions requiring deliberation and reversible experiments where teams receive substantial autonomy provided they articulate clear hypotheses (Bryar & Carr, 2021).

Approaches include hypothesis-based project charters, learning milestone funding, experimental portfolio management, evidence-based pivot processes, and innovation accounting systems.

Conclusion

Organizations navigating today's combinatorial innovation landscape must identify useful possibilities among astronomically large sets of potential combinations. This article has argued that theory-driven experimentation provides an important mechanism, generating combinatorial salience by enabling organizational actors to reason causally and design experiments that deliberately probe promising pathways.

Organizations that cultivate theory-driven innovation capabilities—through structured hypothesis development, systematic experimentation, psychological safety, distributed problem-solving, capability building, learning systems, scientific leadership, and adaptive governance—may achieve meaningful performance improvements. Evidence suggests theory-driven approaches can yield advantages across multiple dimensions.

Yet adoption remains limited because prevailing organizational architectures were designed for execution of predetermined plans rather than rigorous testing of uncertain hypotheses. Transformation requires fundamental reconsideration of how organizations frame decisions, allocate resources, evaluate performance, and develop capabilities.

The path forward involves assessing current practices against theory-driven principles, piloting initiatives demonstrating feasibility, developing embedded capabilities, redesigning governance systems, and consistently modeling scientific reasoning through leadership. By treating employees as capable theorizers and building institutional architectures that enable systematic experimentation, organizations can transform the combinatorial challenge from paralyzing complexity into tractable opportunity.

References

1. Arthur, W. B. (2009). The nature of technology: What it is and how it evolves. Free Press.

2. Brown, T. (2008). Design thinking. Harvard Business Review, 86(6), 84-92.

3. Bryar, C., & Carr, B. (2021). Working backwards: Insights, stories, and secrets from inside Amazon. St. Martin's Press.

4. Camuffo, A., Cordova, A., Gambardella, A., & Spina, C. (2020). A scientific approach to entrepreneurial decision making: Evidence from a randomized control trial. Management Science, 66(2), 564-586.

5. Catmull, E. (2014). Creativity, Inc.: Overcoming the unseen forces that stand in the way of true inspiration. Random House.

6. Cooper, R. G. (2008). Perspective: The Stage-Gate idea-to-launch process—Update, what's new, and NexGen systems. Journal of Product Innovation Management, 25(3), 213-232.

7. Cronin, M. A., & Weingart, L. R. (2007). Representational gaps, information processing, and conflict in functionally diverse teams. Academy of Management Review, 32(3), 761-773.

8. Edmondson, A. C. (1999). Psychological safety and learning behavior in work teams. Administrative Science Quarterly, 44(2), 350-383.

9. Eisenhardt, K. M., & Tabrizi, B. N. (1995). Accelerating adaptive processes: Product innovation in the global computer industry. Administrative Science Quarterly, 40(1), 84-110.

10. Felin, T., & Singell, M. (2025). Technology: Theory-driven experimentation and combinatorial salience. European Economic Review, Special Issue on "Explaining Technology."

11. Fleming, L., & Sorenson, O. (2004). Science as a map in technological search. Strategic Management Journal, 25(8-9), 909-928.

12. Frazier, M. L., Fainshmidt, S., Klinger, R. L., Pezeshkan, A., & Vracheva, V. (2017). Psychological safety: A meta-analytic review and extension. Personnel Psychology, 70(1), 113-165.

13. Garnier, J. P. (2008). Rebuilding the R&D engine in big pharma. Harvard Business Review, 86(5), 68-76.

14. Hong, L., & Page, S. E. (2004). Groups of diverse problem solvers can outperform groups of high-ability problem solvers. Proceedings of the National Academy of Sciences, 101(46), 16385-16389.

15. Huston, L., & Sakkab, N. (2006). Connect and develop: Inside Procter & Gamble's new model for innovation. Harvard Business Review, 84(3), 58-66.

16. Kohavi, R., Tang, D., & Xu, Y. (2020). Trustworthy online controlled experiments: A practical guide to A/B testing. Cambridge University Press.

17. Liker, J. K. (2004). The Toyota way: 14 management principles from the world's greatest manufacturer. McGraw-Hill.

18. Nadella, S. (2017). Hit refresh: The quest to rediscover Microsoft's soul and imagine a better future for everyone. HarperBusiness.

19. Ries, E. (2011). The lean startup: How today's entrepreneurs use continuous innovation to create radically successful businesses. Crown Business.

20. Smith, K. G., Collins, C. J., & Clark, K. D. (2005). Existing knowledge, knowledge creation capability, and the rate of new product introduction in high-technology firms. Academy of Management Journal, 48(2), 346-357.

21. Thomke, S., & Manzi, J. (2014). The discipline of business experimentation. Harvard Business Review, 92(12), 70-79.

22. Weitzman, M. L. (1998). Recombinant growth. The Quarterly Journal of Economics, 113(2), 331-360.

---

Jonathan H. Westover, PhD is Chief Academic & Learning Officer (HCI Academy); Associate Dean and Director of HR Programs (WGU); Professor, Organizational Leadership (UVU); OD/HR/Leadership Consultant (Human Capital Innovations). Read Jonathan Westover's executive profile here.

---

Suggested Citation: Westover, J. H. (2025). Theory-Driven Innovation in Organizations: From Combinatorial Possibilities to Practical Breakthroughs. Human Capital Leadership Review, 27(2). doi.org/10.70175/hclreview.2020.27.2.1