Rethink moonshots – they do not fail. Risk averse research systems do.

May 18, 2026

By Shaun Coffey FTSE FAIA CRSNZ FAICD

Moonshots are often described as high risk, high reward endeavours. In scientific systems, this framing can be misleading. Properly designed moonshot programs may, in fact, reduce systemic risk because they change how failure is understood.

Traditional project funding frequently assumes a linear relationship between investment, activity, and outcome. A project is proposed, milestones are established, and success is measured against predefined deliverables. When the project does not achieve its intended endpoint, it is often labelled a failure. This creates an environment in which researchers become cautious, incremental, and defensive in framing ambition.

Moonshot approaches operate differently. They define success at the level of the long-term challenge rather than at the level of every individual project. The moonshot itself remains stable, but the pathway toward it becomes adaptive.

This distinction is critical to our understanding of failure.

For example, if the objective is drought-tolerant cereals for semi-arid Africa, or nitrogen-fixing cereals, or methane-reducing livestock systems, or radically improved photosynthetic efficiency, it is unreasonable to expect every experiment, platform, or consortium to deliver the final answer. The purpose of individual investments is not always to solve the entire problem directly. Their contribution may instead be to narrow uncertainty, eliminate dead ends, refine hypotheses, generate enabling technologies, or reveal more intelligent research questions.

In this sense, apparent “failure” reflects scientific progress.

A research program that conclusively demonstrates a pathway will not work has still created value. It prevents future resources from being wasted pursuing the same direction. It sharpens strategic focus and improves collective understanding of the system being studied. Equally, a project that develops a new genomic tool, data architecture, phenotyping method, or modelling capability may not achieve the moonshot itself, yet may enable dozens of later discoveries.

Moonshot thinking, therefore, spreads risk differently across the innovation system. Instead of demanding certainty from each project, it accepts uncertainty as intrinsic to transformational science. What matters is whether the overall portfolio advances understanding and capability toward the strategic end point.

This type of thniking is particularly important in international agricultural research where the problems being addressed are biologically complex, environmentally variable, and socially embedded. Step-change advances in food security, climate resilience, soil restoration, or low-emissions agriculture are unlikely to emerge from narrowly bounded three-year projects optimised for immediate reporting cycles. They require persistence, experimentation, and institutional tolerance for ambiguity.

The history of major scientific breakthroughs supports this perspective. Few transformative discoveries emerged through uninterrupted sequences of successful projects. Most involved false starts, competing hypotheses, technological limitations, accidental findings, and repeated reframing of the problem itself. Scientific advancement is rarely linear. It is accumulative and adaptive.

Moonshot programs also create another important benefit. They provide a coherent narrative capable of aligning fragmented research efforts. Researchers from different disciplines, institutions, and countries can contribute distinct pieces of capability toward a shared strategic ambition. One team may improve germplasm. Another may develop sensing technologies. Another may solve a data integration challenge. Another may identify socioeconomic barriers to adoption. Not all contributions look transformational in isolation. Collectively, however, they may create the conditions for transformation.

This reframes how institutions should think about accountability. The question should not always be: “Did this project solve the problem?”

Sometimes the more important questions are:

• What uncertainty did this work reduce?
• What new capability did it create?
• What assumptions did it challenge?
• What pathways did it eliminate?
• What enabling knowledge emerged?
• How did it improve the next generation of research?

Seen this way, moonshots are not reckless exercises in scientific optimism. Properly governed, they are disciplined systems for organising ambitious learning over long time horizons.

The real danger may not be aiming too high. It may be that research systems are constructed so risk-averse, fragmented, and administratively constrained that genuinely transformative advances become structurally impossible. We end up in the weeds, trying to pick winners among a plethora of projects with merit.