I've watched 'pilot' become one of the most useful and most limiting words in almost every organization I work with. Useful because it lowers resistance and makes buy-in easier. Nobody has to fully commit to a pilot. Limiting for exactly the same reason. When the obstacles that would need to change for something to scale never get addressed, and workarounds substitute for real process changes, the pilot ends right where it began.The result has a name, even if most institutions don't use it: pilot purgatory. The work gets done, the feedback is positive, and then everything quietly returns to how it was before. Not because the idea didn't work. Because the conditions for it to work beyond the pilot were never created.What Pilot Purgatory Looks LikeThe pattern tends to show up in recognizable ways. Most pilots are built to answer the question "Can this work?" But that's almost never the right question. With AI tools, the answer is typically yes. The real questions are:<ul><li>What would it take to make this work consistently at scale?</li><li>Who needs to be involved for this to become part of how we operate?</li><li>What changes, in workflows, governance, and culture, need to happen alongside tool implementation?</li><li>How do we measure progress and hold ourselves accountable?</li></ul>Pilots rarely ask these questions because they aren't designed to. They're designed to produce a verdict, not a path forward.There's a second pattern underneath that one. A successful pilot demonstrates capability, not readiness. The tool works, the early users are enthusiastic, and leadership concludes that the hard part is done. In reality, the hard part hasn't started. The pilot proved the idea could work under favorable conditions. It said nothing about whether the organization can support it consistently, at scale, for the people and use cases that weren't in the room. That gap between demonstrated capability and actual readiness is where most pilots stall, and where false confidence quietly leads to under-resourced rollouts and aggressive timelines that the work can't carry.The Problem with Low StakesThere's another issue with the pilot framing: it signals low stakes. When something is called a pilot everyone involved understands implicitly that it might not go anywhere. That understanding shapes behavior.People invest differently in things they believe will make a difference. When a project is positioned as tentative, commitment tends to be tentative too. So does engagement. The project becomes a test of the concept rather than a genuine attempt to create change.You can't accurately test whether something will work at scale or sustain itself over time but that's exactly what a pilot design invites you to do.What the Pilot Didn't Need (But Real Adoption Will)Part of what makes pilots feel successful is that they don't have to carry the weight of real operations. The pilot ran with a small group of willing participants, a project lead personally available to answer questions, no formal governance review, no accessibility audit, no integration with the systems people actually use, and no plan for what happens when the next round of staff or students arrives.None of that is a flaw in the pilot. It's the point of a pilot. But it explains why so many pilots that "worked" never go anywhere. The moment you try to extend the work beyond the original group, every one of those informal supports becomes a formal requirement. Governance, accessibility, integration, onboarding, and ongoing support all have to be designed, resourced, and owned. Pilots almost never plan for that, because they were never asked to.What a Catalyst Project Is InsteadWhen we work with organizations on AI transformation we use a different framing: Catalyst Projects.A Catalyst Project is a forward-deployed initiative designed to solve a real institutional problem or capture a real opportunity, with visible results and a built-in plan for what happens next. The difference from a pilot isn't just semantic. It's structural.Catalyst Projects are designed for scale from the start.Before we begin, we've identified what success looks like, what would need to be true for this to expand beyond its initial scope, and who would need to be involved in that expansion. The end of the project isn't a verdict. It's a launch point.Catalyst Projects are scoped to matter.We identify what we call Green Light Zones: use cases that are worth doing and positioned to succeed, meaningful enough that results will be noticed, and structured in ways that make scaling a natural next step.Catalyst Projects create evidence, not just experience.Because we measure outcomes from the start, using our AI Capacity Index, we're building a record of what changed and by how much. That evidence is what makes internal advocacy possible. It's what counteracts skepticism and builds the case for moving forward. Experience produces anecdotes. Evidence produces momentum.Catalyst Projects build internal capacity, not just results.Every Catalyst Project is both a solution and a learning investment in the people, practices, and institutional knowledge that sustain transformation over time.One way to think about it: the tool gets you part of the way. The organizational design and the capacity to operate it get you the rest of the way. Most pilots focus almost entirely on the first half and assume the second half will sort itself out. It doesn't. A Catalyst Project is built around the second half from the beginning, because that's where transformation actually lives.A Different Starting QuestionThe shift from pilot to Catalyst Project starts with a different question. Instead of "What can we test?", the question becomes: "What problem are we actually trying to solve and what would a real solution require?"That question leads to a different kind of project: one with real stakes, real investment, real measurement, and a real plan for what comes next.If your organization has run pilots that went nowhere, you're not alone. The good news is that the problem isn't the technology, the people, or the idea. It's the design. And design can be changed.Most organizations we talk to have at least one idea that's been waiting for the right conditions to move forward. A Catalyst Project might be exactly that.<hr>About the AuthorMarcy Drummond is Managing Director of Eduvators LLC, a Strategic Implementation Partner for AI transformation. She has led AI integration work across higher education, K–12, government, and business settings, and brings direct experience from leadership, operations, and business roles to her work with organizations navigating transformation.<a target="_blank" href="mailto:marcy@eduvators.net">marcy@eduvators.net</a>

AI pilots rarely fail because the idea doesn't work. They fail because they aren't designed to scale. Here's why we use Catalyst Projects instead, and what that difference means in practice.

Why ‘Pilot’ Is a Four-Letter Word: The Case for Catalyst Projects

Read articles, tips, and resources to help you learn, grow, and succeed. Explore our latest posts covering education, community, and personal growth

Blog – Insights, Tips & Resources for Learning and Growth

Explore articles, tips, and insights to help you succeed in learning, community building, and personal development.

Why ‘Pilot’ Is a Four-Letter Word: The Case for Catalyst Projects

Why ‘Pilot’ Is a Four-Letter Word: The Case for Catalyst Projects

Related Posts

New Course: AI-Assisted Grant Research

Capability Erosion: What's Missing from Most Responsible AI Discussions

The AI Implementation Gap: Why Tools and Training Aren’t Enough

New Course: AI-Assisted Grant Writing

New Resource: The Environmental Impact of AI in Higher Education