The executive boardroom buzzed with discussion, a familiar scene unfolding within large organizations grappling with the transformative potential of artificial intelligence. The CEO articulated a vision of AI as a new engine of growth, promising unprecedented expansion. The Chief Information Officer, focused on immediate operational gains, championed the deployment of AI-powered copilots to enhance productivity. Meanwhile, the Chief Operating Officer identified opportunities for widespread process automation, envisioning streamlined workflows and cost efficiencies. The Head of Product saw AI as a catalyst for developing next-generation, intelligent offerings, while the Chief Human Resources Officer voiced concerns about the profound impact AI would have on the workforce, necessitating significant reskilling and restructuring. Despite the diverse perspectives, a consensus emerged: AI represented the future, and innovation was paramount. Yet, as the meeting concluded, each executive departed with a distinct interpretation of the path forward, highlighting a fundamental disconnect in how "innovation" itself was understood.
This disconnect, as observed by seasoned innovation strategist Jeff DeGraff, has become a pervasive "innovation problem" within large enterprises. The single word "innovation" is increasingly used as an umbrella term to describe a wide spectrum of distinct activities: the adoption of new technology, the refinement of existing processes, the launch of novel products, the creation of entirely new business models, the fundamental transformation of an organization, and even reactive responses to crises. This linguistic ambiguity breeds a false sense of alignment, where differing expectations and approaches lead to scattered resources, conflicting priorities, and ultimately, initiatives managed with inappropriate methodologies and measured against irrelevant benchmarks. DeGraff, who has spent nearly four decades advising Fortune 500 companies and governmental bodies, argues that many organizations do not lack innovation; they lack a clear and shared language to define and pursue it.
The Genesis of a Term and its Perilous Ambiguity
DeGraff’s journey into understanding the nuances of innovation began decades ago, during the explosive growth of Domino’s Pizza in the 1980s. As Vice President of New Ventures, he was privy to the company’s ambitious expansion strategies. A retreat with the renowned business author Tom Peters led to DeGraff being affectionately dubbed the "Dean of Innovation," a title that, initially a lighthearted jest, came to represent a lifelong exploration of what truly drives organizational change and growth. His extensive work with a broad range of clients, from global corporations to military strategists and cultural institutions, has solidified his conviction: the word "innovation" is rarely understood uniformly. This lack of shared meaning, he contends, is a primary driver of predictable outcomes: resources are diluted, expectations clash, and initiatives are ill-equipped to succeed because they are managed and evaluated with the wrong frameworks.
The core of the problem, DeGraff suggests, lies in the simplistic framing of innovation. For too long, executives and academics have attempted to categorize innovation into binary choices: incremental versus radical, sustaining versus disruptive, product versus process. While these distinctions offer some clarity, they fail to capture the contextual nature of innovation. The question, he proposes, should not be "Is it innovative?" but rather, "Innovative compared with what?" Innovation, in essence, is a "positive deviance from the norm that creates value." What might be a groundbreaking departure for a legacy organization could be standard practice for a nascent startup. A traditional company implementing AI to revamp a core process might represent a significant leap from its established norms. In contrast, an AI-native startup employing similar technology might simply be adopting industry best practices. The technology itself may be identical, but the starting points and the perceived deviation from the existing landscape are vastly different. This underscores the need to ask: "What norm are we breaking, and what new value are we creating, and for whom?" Without clear answers to these questions, "innovation" remains a vague corporate aspiration.
The Nvidia Case Study: A Paradigm of Value Creation Through Norm Deviation
The meteoric rise of companies like Nvidia serves as a compelling illustration of DeGraff’s thesis. Attributing Nvidia’s success solely to "innovation" is an oversimplification. The company’s strategic brilliance lay in its ability to repeatedly extend its graphics processing capabilities into new and emerging arenas, most notably artificial intelligence computing. This strategic evolution represents the true work of innovation: a deliberate departure from established norms that unlocks and creates new forms of value. As an idea deviates further from the existing paradigm, the challenge of demonstrating its viability and potential impact intensifies. This inherent difficulty is often misunderstood, leading to a reluctance to embrace truly transformative initiatives.
The Data Trap: A Hindrance to True Novelty
A significant impediment to fostering genuine innovation is the inherent organizational inclination towards data-driven decision-making. While essential for managing established operations, this reliance on historical data can stifle novel ideas. The paradox is stark: the more genuinely novel an idea, the less reliable past data becomes. A modest improvement to an existing product can often be modeled with relative accuracy. Customers are known entities, costs can be estimated, and competitors can be studied. However, what happens when a company ventures into an uncharted market or deploys a technology whose capabilities and economic viability are in constant flux?
The advent of generative AI provides a contemporary example. In late 2022, few executives could present a credible five-year return on investment model for generative AI initiatives. The technology itself, associated costs, competitive landscape, regulatory frameworks, and potential use cases were evolving at an unprecedented pace. Organizations that hesitated, waiting for greater certainty, did not necessarily mitigate risk; they simply slowed their learning process. This is the essence of the "data trap." Initiatives with the strongest supporting evidence are often those most closely aligned with what the organization already does. An overemphasis on demanding proof inadvertently favors the familiar, transforming innovation into mere optimization.
The critical shift in thinking, DeGraff advocates, is to move away from asking, "Can you prove this will work?" towards, "What is the cheapest, fastest experiment that will tell us something important?" In mature operations, data informs action. In the realm of innovation, action often precedes and generates the data. Experiments do not serve to validate a pre-existing plan; rather, they are the crucible in which knowledge is forged, forming the foundation upon which a plan can eventually be constructed. This iterative process of learning through doing is fundamental to navigating uncertainty.
Innovation: Not a Monolithic Endeavor
Another common organizational misstep is treating innovation as a singular, uniform activity. In reality, it encompasses a diverse range of challenges, each requiring a distinct management approach. Consider UPS’s continuous efforts to optimize delivery routes. This is primarily an optimization challenge, operating within an established system, with abundant data and measurable progress. Contrast this with a company responding to a sophisticated cyberattack. Here, speed is paramount, often eclipsing the need for perfect information. A legacy corporation venturing into AI-enabled services faces a different set of uncertainties related to customer adoption, pricing strategies, and business models, necessitating small, staged experiments and investments.
Even larger in scope is the transformation underway in the automotive industry, as manufacturers shift towards electric and software-defined vehicles. This represents a monumental undertaking, impacting manufacturing processes, supply chains, talent acquisition, capital allocation, and corporate identity. All these scenarios demand effective management, but they do not require the same management. Yet, many companies subject these disparate initiatives to identical stage-gate processes, funding mechanisms, and performance metrics.
Before selecting a management process, leaders must critically assess:
- How significant is the departure from existing knowledge and capabilities? An incremental improvement might involve familiar customers, technologies, and economic models. A new business model, however, may call all these into question.
- How rapidly must we respond? A crisis demands immediate action, while the maturation of a new business model may take years.
The answers to these questions should dictate the composition of the team, the funding allocated, the governance structure, the metrics employed, and the pace of development, rather than the other way around.
The Misattribution of Change as Innovation
A prevalent issue is the tendency for companies to label almost every form of change as "innovation." The installation of a new Enterprise Resource Planning (ERP) system, the adoption of AI tools like Microsoft Copilot, the reorganization of a business unit, the digitization of a process, or the hosting of a hackathon—while potentially valuable—are not inherently innovative in the same way. A useful distinction lies in understanding the roles of creativity, innovation, and change. Creativity generates possibilities. Innovation develops and rigorously tests novel possibilities to create tangible value. Change management focuses on enabling the adoption and scaling of those validated innovations.
Purchasing an AI platform is a form of technology acquisition. Successfully integrating it and achieving widespread employee adoption is change management. However, discovering a fundamentally new way to create customer value through that AI platform—that is innovation. A company can excel at the first two without necessarily achieving the third. Consequently, merely counting AI pilot projects provides limited insight into an organization’s genuine innovative capacity. Automating an existing process may boost productivity, and deploying copilots might accelerate individual work. These can be excellent investments, but adoption is not invention, and efficiency gains do not automatically equate to innovation.
The Peril of False Alignment
Returning to the executive meeting scenario, the divergence in objectives becomes clear. The CFO might be focused on cost reduction within a 12-month timeframe, while the Head of Product aims for new revenue streams. The CIO prioritizes a secure and robust technology platform, and the CHRO seeks to redesign work processes. All these objectives may be valid, but without explicit acknowledgment and articulation of these differing aims, the initiative will inevitably be judged against conflicting expectations. The organization may not have an execution problem; it may have a fundamental language problem.
Leaders often attempt to force alignment in uncertain situations. However, disagreement in such contexts is valuable information. One executive might perceive an efficiency opportunity, another a competitive threat, and a third a nascent business model. The goal should not be to prematurely eliminate these differing viewpoints. Instead, the aim should be to make these distinctions visible. This is not dysfunction; it is clarity.
A Five-Question Framework for Navigating Innovation
To address this pervasive ambiguity, DeGraff proposes a pragmatic five-question framework that leaders can employ in meetings, funding reviews, and everyday conversations. This framework moves beyond abstract definitions and provides actionable prompts:
1. What norm are we breaking?
This question compels a clear identification of the existing standard being challenged. Is it a norm within the company, the industry, the technological landscape, or customer expectations? If a leader cannot articulate the norm being broken, the initiative might be a mere improvement of existing practices rather than a genuine innovation.
2. What new value are we creating, and for whom?
Innovation is not about novelty for its own sake. A statement like, "We need an AI strategy," is as unhelpful as stating, "We need an internet strategy" would have been two decades ago. The technology itself is not the strategy; the strategy lies in what becomes possible that was not previously achievable. If the value proposition cannot be clearly articulated, the initiative risks becoming mere "technology theater."
3. How big is the departure from what we already know?
This question helps gauge the level of uncertainty and risk involved. An incremental improvement typically involves familiar customers, technologies, and economic assumptions. A disruptive new business model, conversely, might call all three into question. Small departures can often be planned with greater certainty, while large departures necessitate a learning-oriented approach. The size of the investment should be matched to the availability of knowledge.
4. How fast do we need to move?
Speed is not an inherent virtue in all situations. A cybersecurity breach demands immediate action, measured in hours. The development and maturation of a new business model, however, may take years. The critical question is not merely, "How can we accelerate?" but rather, "What is the appropriate speed for this specific context of uncertainty?"
5. What is the next experiment?
This question shifts the focus from long-term planning to immediate, actionable learning. Instead of asking for a five-year plan or a definitive scaling timeline, the emphasis is on identifying the most crucial question to answer next. A well-designed experiment should address a significant unknown at an acceptable cost. Will customers adopt this? Can the technology function under real-world conditions? Will someone pay for it? What underlying assumption, if proven false, would lead to the initiative’s failure? The objective is not to validate the team’s initial hypothesis but to acquire genuine knowledge. The ultimate failure lies in investing years in developing something that could have been disproven in a matter of weeks.
Transforming the Conversation, Navigating the Future
By shifting the questions, organizations can fundamentally transform their conversations around innovation. Instead of asking, "Is it innovative?" leaders should inquire, "What norm are we breaking?" Instead of demanding an ROI, the question should be, "What must we learn before investing more?" Rather than asking, "What is the plan?" the focus should be on, "What is the next experiment?" And in place of "Are we aligned?" the more productive inquiry is, "Where do we see the problem differently?" Finally, instead of "How quickly can we scale?" the relevant question is, "Have we created enough value to warrant scaling yet?"
These revised questions do not eliminate uncertainty; they render it manageable. An optimization effort should be accountable for tangible results. An experiment should be accountable for learning. A crisis response is accountable for speed. A transformation is accountable for building new organizational capabilities. When leaders employ a singular language and set of metrics for these disparate activities, confusion is inevitable. Differentiating among them, however, enables more intelligent and effective management.
As artificial intelligence accelerates the pace of technological change, an increasing number of initiatives will be labeled as innovative. More companies will launch pilots, establish labs, and embark on strategic ventures. Significant financial resources will be committed before leaders reach a shared understanding of the problem they are truly attempting to solve. The organizations that ultimately succeed will not necessarily be those with the greatest number of ideas, but rather those capable of discerning between optimization and invention, adoption and experimentation, evidence and assumptions, and plans and learning. They will possess the wisdom to know when to demand data and when to create it, when to move with urgency and when to exercise patience, and when to scale and when to continue experimenting.
Innovation, by its very nature, is not a static entity. It evolves with the organization, the industry, the technology, and the prevailing economic and social climate. The pursuit of a singular, perfect definition of innovation has therefore always been a futile endeavor. What organizations truly need is not another abstract definition, but a practical, usable language that leaders can effectively employ to navigate the complex and ever-changing landscape of innovation.
