By Andrew Maxwell

Dr. Andrew Maxwell is the Bergeron Chair in Technology Entrepreneurship in the Lassonde School of Engineering at York University.              

Innovation is not a one-time event or the result of inspiration alone. It is inherently uncertain and prone to failure.

As Woody Allen quipped, “If you're not failing every now and again, it's a sign you're not doing anything very innovative.”

This humorous truth encapsulates a central dilemma of innovation policy: if we are genuinely innovating – pursuing new ideas, technologies or business models – then some level of failure is inevitable. Without recognizing this, we risk building policy frameworks that punish experimentation and stifle the very innovation we hope to encourage.

Yet failure poses a profound problem for political systems. Politicians are rewarded for declaring success, delivering results and avoiding controversy. As such, acknowledging the possibility – or necessity – of failure runs counter to political instincts.

In the public sphere, programs that fall short of their goals are often quietly shelved rather than openly examined. This suppresses the very feedback we need to learn and improve. The fear of public scrutiny often leads to a deeper problem: failure is not only hidden – it is repeated.

Lessons go unlearned, and institutional memory is eroded with each political cycle. Moreover, when we fail to acknowledge the risk of failure up front, we fail to design for it.  Innovation efforts become performative, framed as guaranteed successes.

In reality, when setbacks occur, they are often ignored or denied, leading to the continuation of projects that no longer merit support. This is the sunk cost fallacy at work – sustained by politics, pride and institutional inertia.

To address this, we must embrace the reality that innovation is a process – a complex, staged sequence of activities involving exploration, validation, development and eventual implementation, each carrying its own risks and requiring its own metrics. In this context, applying a quality management approach to the innovation process is not just prudent – it is essential.

As outlined in the foundational work of U.S. engineer and management consultant Joseph Juran and U.S. business theorist and management consultant W. Edwards Deming, quality management emphasizes not only reducing defects, but continuously improving processes through feedback, measurement and accountability.

Just as quality assurance in manufacturing reduces variation and improves outcomes, a similar approach to innovation allows us to track where things go wrong, understand why, and refine how we operate. This approach also calls for rigorous process mapping to visualize how innovation flows through various stages, to identify misalignments, bottlenecks or poor handoffs. When the output of one stage does not match the requirements of the next, innovations can stall or fail unnecessarily. These disconnects are especially common when innovations move between actors – such as from academic researchers to commercial partners or from government programs to end users.

Canada needs an independent organization to assess innovation

Canada’s current innovation strategy suffers from fragmentation, short-termism and a lack of coherence. Programs are often designed in silos, aligned more with ministerial mandates or regional equity goals than with a unified national strategy for economic transformation.

While political leaders may intend to stimulate innovation, the result is a patchwork of initiatives that fail to build cumulative advantage. There is often little coordination across departments, little continuity across political cycles, and little attention to where Canada has the potential for true global leadership.

These programs are also frequently reactive, rushed to meet political timelines, and framed to appear fair rather than to be effective. Funding calls are launched with short notice, with success measured by geographic or demographic distribution rather than strategic impact.

While equity is essential, equating fairness with spreading resources thinly across all sectors and regions can undermine the very focus needed to build strong clusters or support ambitious national missions.

Rather than concentrating on areas where Canada has demonstrable strength – such as AI, clean energy or advanced manufacturing – innovation efforts are often diluted by the desire to appease all stakeholders equally. As a result, we rarely invest at the scale or duration necessary to move the needle.

To support this system-wide improvement, Canada needs an organization that is independent of program delivery but empowered to observe, assess and recommend changes. Its mission should be to identify where innovation processes break down and propose practical, evidence-based improvements.

For this model to work, accurate and comprehensive data is essential. We must move beyond political statements or anecdotal case studies and instead collect real performance data, tracking outcomes, stage-by-stage progress and decision-making points.

Furthermore, we must shift from measuring inputs (such as dollars spent, patents filed or startups launched) to measuring outcomes – such as adoption, value creation, productivity impact and learning.

Only by focusing on what emerges from the process, rather than what goes into it, can we identify underperforming initiatives, correct structural weaknesses and improve our capacity to innovate over time.

This challenge becomes even more pronounced when multiple actors are involved across the innovation value chain. A striking example lies in university-based research. Researchers are often incentivized by publication metrics rather than the creation of user value. As a result, innovations that emerge from academic settings may be technically promising but lack the market fit or usability to be adopted.

This disconnect – between upstream knowledge creation and downstream application – is a major source of stalled technologies – breakthroughs that remain confined to academic journals instead of reaching users who could benefit from them.

We should learn from organizations like the Fraser Institute and C.D. Howe Institute, which focus on performance measurement and policy effectiveness in education and economic governance. Their work underscores the importance of outcome-focused metrics and longitudinal analysis – principles that could equally transform how we govern innovation.

While this article places less emphasis on institutional proposals, the lessons from these examples show the power of independent, data-informed critique. These models demonstrate how credible, non-partisan analysis can guide better decision-making without being constrained by short-term political cycles.

If Canada is to become a truly innovative country, we must stop treating failure as a political liability and start seeing it as a diagnostic tool. The only unacceptable failure is the one we refuse to learn from.

R$