Bahcall is a physicist and biotech entrepreneur, which is a rarer combination than it sounds, and it gives the book an odd credibility — theoretical precision paired with someone who’s actually had to ship a product. The core argument borrows a metaphor from physics: phase transitions. Water doesn’t cool gradually into ice. It sits there as liquid, liquid, liquid, and then at a specific threshold it snaps into a different state entirely. Organizations, Bahcall argues, do the same thing. They flip between a state where people are rewarded for chasing risky, unconventional ideas (“loonshots”) and a state where people are rewarded for defending what already works (“franchises”). The flip happens at specific, measurable organizational parameters — which means, in theory, a leader can find the dial and turn it.
The book splits its time between the theory — building out the physics metaphor and working through its implications — and a run of case studies that put the theory through its paces. These are some of the more engaging case studies in recent business writing: pre-WWII radar, the discovery of the first effective cancer drugs, the rise and slow death of Pan American Airways, the early histories of Apple and Disney. Same pattern every time. The organizations that produced the breakthrough had, somehow, held onto what Bahcall calls the “magic number” — the right ratio between loonshot nurseries and franchise forces.
The Phase Transition Model
The physics isn’t decoration. Bahcall shows the actual math of phase transitions — a critical parameter that decides which state a system lands in — maps onto organizational behavior with surprising fidelity. Temperature decides whether water is ice, liquid, or vapor. In an organization, a structural parameter Bahcall calls M — roughly, the number of management layers — decides whether the place is in loonshot mode or franchise mode.
Here’s the mechanism. As organizations grow and stack management layers, individual incentives shift underneath everyone’s feet, usually without anyone noticing it happening. In a small org, everyone knows everyone, information moves fast, and doing excellent work on a real problem is the fastest way up. In a large org, advancement runs through managers making judgment calls on limited information — how you’re positioned politically, how well you manage upward, how visible your metrics are to someone without deep domain knowledge. In that environment the rational move for a talented person isn’t the risky unconventional idea. Too likely to fail. Too easily misread by a manager without the expertise to evaluate it. The rational move is the safe, franchise-defending project where success is likely and easy to explain in a status update.
And the shift from loonshot-nurturing to franchise-defending doesn’t creep. It snaps — at a specific size, a specific structure — the same way water doesn’t slowly become ice but flips over at 32 degrees. This is what makes Bahcall’s model sharper than the usual “maintain an innovation culture” advice. It isn’t about mindset, though mindset matters some. It’s about structural parameters, and structural parameters can be measured and, with effort, managed.
The Historical Cases: Radar and the Death of Innovation
The most gripping of the historical cases is radar in World War Two — the same invention, developed and deployed brilliantly in one place, developed and left to rot in a drawer in another, depending entirely on the receiving institution’s structure. Radar was built in the U.S. in the 1930s at the National Bureau of Standards by a small research team. It worked. The Navy tested it. It worked in the tests too. And then nothing happened. The technology sat in a file while the fleet went unequipped. On December 7, 1941, the Army actually had a radar station running on Oahu, and it picked up the incoming Japanese planes over an hour before the attack hit. The operators weren’t sure what they were looking at. They called their supervisor. He waved it off. The attack proceeded exactly as planned by the other side.
Why would a military sit on a technology that worked, that had been tested, that could have prevented — or at minimum badly blunted — the worst military disaster in American history? Bahcall’s answer is structural, not personal. The late-1930s U.S. military was a franchise-defending organization through and through. Advancement meant navigating the existing hierarchy, managing relationships with senior officers, demonstrating competence in the skills that hierarchy already recognized. The engineers who understood radar had neither the institutional standing nor the vocabulary to reach the officers who controlled procurement. And the officers weren’t stupid, whatever the tidy version of this story wants to claim. They were responding, rationally, to an incentive system that rewarded conventional excellence and punished disruption.
Compare that to how the British handled the same technology. The RAF built a direct structural line between the scientists developing radar and the officers commanding air defense — no layers, no filter. Same rough technology as the Americans, same rough timeline. Different structure, wildly different outcome: the British deployed radar effectively through the entire Battle of Britain, and Germany’s failure to break the RAF that summer traces substantially back to the advantage that one structural decision bought them.
The Cancer Drug Discovery Cases
Bahcall’s background is biotech, and the book’s most technically grounded cases come from cancer drug development — a history that is, more than most people realize, a history of breakthroughs that nearly didn’t happen. Compounds with real efficacy got shelved by big pharma because they didn’t fit the existing research pipeline, because the wrong person championed them, or because the company receiving them happened to be sitting deep in its franchise-defending phase at exactly the wrong moment.
The sharpest example is imatinib (Gleevec) — the drug that turned chronic myeloid leukemia from a near-certain death sentence into a disease with a 90% survival rate. Getting there took years of a small group of Novartis researchers pushing against their own management, which saw the compound as commercially unviable given how small the patient population looked on paper. The researchers kept pushing. The clinical results turned out extraordinary. The drug got approved eventually. But those years — the years a potentially life-saving compound spent quietly dying in a corporate drawer — are exactly the failure mode Bahcall’s model is built to diagnose.
The Moses Trap and the Innovation Killer
The name is biblical, obviously — a leader so central, so visionary, that the whole operation depended on him, which meant that the moment he was gone, nothing worked. Bahcall traces the pattern through Steve Jobs’s second run at Apple and George Lucas at Lucasfilm — both men who built brilliant loonshot cultures early on and then, later, became the very bottleneck preventing anyone else from developing a new loonshot.
The antidote is what Bahcall calls “phase separation” — a deliberate structural split between loonshot development and franchise management, with clear rules governing how information and ideas move between the two, and authority structures that shield each phase from the other’s incentives. This is the structural version of “innovation culture,” stripped of the vague ethos and replaced with an actual design: separate the people doing early exploratory work from the people running the existing successful products, and build the right connective tissue between them.
The Bush-Vail Rules
Bahcall pulls four rules out of his study of history’s most successful innovation cultures and names them the “Bush-Vail Rules,” after Vannevar Bush (the science administrator who built the research infrastructure behind radar and the atomic bomb) and Theodore Vail (who turned AT&T from a phone company into one of history’s most innovative research organizations by founding Bell Labs). The rules: separate your loonshots from your franchises; build a dynamic equilibrium between the two; spread a system mindset; raise the magic number.
That fourth rule — raise the magic number — is the structural lever the physics model actually gives you. The magic number is the parameter that decides which phase the organization is in. By identifying the specific structural knobs that determine whether people chase loonshots or defend franchises — span of control, the ratio of equity to salary, the criteria used in performance review — leaders can push those knobs toward the loonshot-nurturing side. That’s not soft culture-talk. It’s hard structural design, and that specificity is what makes the model actually usable, unlike most innovation-culture advice.
The Limits of the Model
The physics metaphor is powerful, not perfect, and an honest reader should hold both at once. The mapping from physical phase transitions to organizational dynamics is suggestive, not exact — organizations aren’t physical systems, and thermodynamic math doesn’t apply to them literally. The “magic number” isn’t as cleanly calculable as Bahcall sometimes implies, and the structural fixes he recommends need a lot of contextual judgment to actually implement.
The case studies, engaging as they are, tend to be cases that fit the model neatly rather than cases that stress-test it. The pharma examples in particular benefit from — and are shaped by — Bahcall’s insider vantage point. A more skeptical pass might turn up cases where the phase transition model predicts the wrong outcome, or gets the mechanism wrong entirely.
Real limitations. They don’t undercut the book’s core value, though. The central claim — that organizations reliably suppress their most important innovations through structural incentives favoring franchise defense over loonshot pursuit, and that specific structural moves can fix it — is grounded and useful. More precise than the standard innovation-culture line. More actionable than the motivational stuff. More tied to actual organizational history than most management theory bothers to be.
What Loonshots Contributes
That specificity is the whole selling point, and it’s why the book belongs in a different category than most of what shares shelf space with it. Bahcall is telling you something you can actually do: measure your organization’s structural parameters, find where they’re choking loonshot development, adjust them. The adjustments still require judgment and context — nobody’s handing you a formula. But they’re adjustments to concrete structural features, not appeals to the vague “culture” most management books gesture at without ever pinning down.
And on top of the theory, it’s just a good read. The historical cases pull you along, the prose is clean, the physics gets explained with enough care to illuminate without scaring off readers who never took a physics class. Bahcall has a real knack for making complicated ideas legible without stripping out the complexity that makes them useful in the first place, and Loonshots is among the better examples of popular-science writing turned on organizational problems. Anyone who’s watched a good idea die inside a large organization and wondered why gets both an answer and something resembling a way forward here.
The Disney and Pan Am Cases
Two of the more illuminating cases involve organizations that produced extraordinary loonshots and then simply lost the capacity to keep producing them. Walt Disney’s studio in the 1930s and ’40s put out a run of revolutionary animated features — Snow White, Fantasia, Bambi, Dumbo — that reshaped the art form and put the studio at the top of animation. Then, through the ’50s and ’60s, the output shifted hard toward safer, more formulaic work. The studio that had invented the medium was now turning out competent, forgettable stuff. What happened?
Bahcall points to the structural changes that came with success. As the studio grew, the management hierarchy expanded, the successful formulas calcified, and the people who’d made the original loonshots — who took the risk on Snow White when everyone told them a feature-length animated film was impossible — got replaced by managers who’d learned to optimize what success had already established. The nursery that produced the original breakthroughs gave way to a franchise-defending machine that could execute known approaches well and generate new ones not at all.
Pan American Airways tells a parallel story from a different industry. Pan Am was, for a long stretch, one of the most innovative companies in 20th-century American business — it pioneered transoceanic commercial flight, brought the Boeing 707 into commercial service, launched the 747 program, built the reservation system that became the industry template. Thirty years, from the mid-1930s to the late ’60s, of being the most technologically aggressive airline on earth, constantly commissioning new aircraft that changed what flying meant. Then the engine just stalled. The company kept operating as a major carrier for another two decades before going bankrupt in 1991, but nothing in those last twenty years came close to what it had done in the first thirty.
Same structural story as Disney: growth, more hierarchy, entrenched formulas, the slow drift from loonshot-nurturing to franchise-defending. The people who built the original innovations stopped driving strategy; the people who’d learned to manage the existing operation took over. The company that basically invented the modern airline couldn’t adapt to the wave of innovation that eventually swallowed it. Bahcall uses both cases to make the point that no organization is immune to this phase transition, and that only deliberate structural intervention stops it.
Applying the Framework Today
The practical question Loonshots puts to anyone who runs — or hopes to run — an organization: where does yours sit on the curve? Are people mostly motivated to chase risky, unconventional ideas, or mostly motivated to defend and polish what’s already working? Probably both, honestly, in different corners of the organization, and the real question is whether the balance is right and whether the parameters can be nudged.
Bahcall offers a handful of diagnostic questions. Do the people pursuing your most important new ideas get rewarded when those ideas succeed — or does the credit flow mainly to the managers who signed off? Do failed loonshots get treated as evidence of incompetence and punished, or as valuable information and treated accordingly? Do the people developing early-stage ideas have a direct line to the decision-makers who can greenlight development, or are they buried under management layers that filter and distort everything before it reaches the top?
These are structural questions, not cultural ones, and that’s the point — structure sets incentives, incentives set behavior. An organization that wants loonshots to survive has to make pursuing one a rational career move, not a form of idealistic self-sacrifice. Reward loonshot development structurally, and talented people will chase loonshots. Penalize it structurally, and they won’t — no matter how many “innovation culture” posters go up on the walls.
The Deeper Lesson
The deepest lesson in Loonshots might be about the relationship between success and vulnerability. The organizations most at risk of losing their loonshot culture are, perversely, the ones that already succeeded — the ones that produced the transformative innovation that became the franchise they now spend all their energy defending. Success grows the organization, and growth shifts the phase-transition parameter. Success entrenches the formulas that franchise defense protects. Success builds the career incentives that reward polishing over exploring. The very win worth celebrating plants the seed of the failure that follows it.
A sobering thought, and a practically important one. It means the leaders of successful organizations need to be most alert to their loonshot culture at the exact moment they feel least worried about it. The time to put structural protections in place isn’t during a crisis — it’s while the organization is winning, while the franchise is thriving, before the management layers have piled up and the phase-transition parameter has crept past the threshold. Like most preventive medicine: the right time to act is before the symptom shows up.
Bahcall’s book is both diagnosis and prescription — a clear account of why organizations fail to protect their most important innovations, and a specific set of structural moves that can stop that failure. The diagnosis leans on physics and history. The prescription leans on organizations that actually held the loonshot-nurturing phase over long stretches. Both hold up, and together they’re one of the more original contributions to management writing in years. For anyone who cares about building organizations capable of real innovation — not the performance of it, the actual thing — Loonshots earns its place on the required list.
The Statistical Mechanic’s Lens
What separates Bahcall from most management thinking is that he looks at organizations as statistical systems, not as collections of individually motivated people. The standard advice fixates on the individual: hire the right people, give them the right incentives, build the right culture, and good outcomes follow. Bahcall’s point cuts against that — individual behavior is substantially shaped by structural parameters that apply to everyone in the system at once, and shifting those parameters produces predictable shifts in aggregate behavior no matter who the specific individuals are.
That’s the physicist’s lens turned on organizational dynamics, and it lands on conclusions that feel backwards from the conventional management view. Conventional view: not innovating? Go find better innovators. Bahcall’s view: not innovating? Your structural parameters are sitting in the franchise-defending phase, and swapping out personnel won’t fix it unless the structure changes too. The problem was never the people. It’s the incentive terrain the structure builds, and that terrain applies to everyone standing on it the same way.
None of this means individual talent doesn’t matter — Bahcall wouldn’t claim that. But a talented person’s output is substantially shaped by the structural environment they’re standing in. A talented loonshot developer stuck in a franchise-defending organization learns, correctly, not to bother developing loonshots, because the structure won’t reward it. The same person in a loonshot-nurturing organization develops loonshots freely, because the structure does. The output gap between the two organizations isn’t a talent gap. It’s a structural one.
Genuinely powerful, and it’s backed by a more rigorous analytical framework than most management books bother building. Loonshots isn’t the easiest business book around — the physics needs some patience, and Bahcall’s commitment to rigor occasionally makes something complicated that could’ve stayed simple. But the complexity earns its keep. It buys a level of precision most management thinking never reaches, and that precision is what turns the prescriptions actionable instead of merely inspirational. Anyone who wants to know not just that their organization should support innovation, but why it currently doesn’t and what to specifically do about it, will find the most rigorous answer available in the popular management shelf right here.
Conclusion: A New Framework for Old Problems
Safi Bahcall wrote a book worth reading by anyone trying to understand why organizations fail to protect their most important innovations, and what to do about it. The physics-based framework beats the usual culture-change prescriptions on precision. The historical cases are more carefully worked than the anecdotes that fill most business books. The prescriptions are sharper than the standard exhortations to “embrace failure” and “think outside the box” that saturate the innovation literature.
The deepest contribution here is reframing innovation failure as a structural problem, not a talent problem, not a culture problem. Organizations don’t fail to innovate for lack of talented people. They fail because their structure makes franchise defense more rational than loonshot pursuit for the talented people they already have. The fix is structural: redesign the incentive terrain by adjusting Bahcall’s specific parameters, separate loonshot development from franchise management, build the connections that let loonshots graduate into franchises when the time’s right, and hold the dynamic equilibrium that lets both phases operate at full strength.
Not simple. If it were, every organization would already be doing it. But it’s specific enough to attempt, rigorous enough to evaluate, and grounded in enough history to be more than management-consulting fashion. Loonshots is a serious book about a serious problem, written with real scientific rigor and genuine literary craft. It belongs on the shelf of anyone who leads an organization, manages a team, or spends any time wondering why some institutions keep their capacity for transformative innovation while most eventually lose it.
Worth remembering too: the history of innovation isn’t a smooth arc of inevitable technological progress. It’s contingent, messy, full of near-misses where transformative ideas almost died before they got the chance to change anything. Radar nearly went undeployed. Imatinib nearly stayed in a drawer. The internet’s foundational protocols nearly went unpublished. The World Wide Web nearly stayed a CERN side project. Every time, the gap between the world-changing outcome and the wasted shot wasn’t the quality of the idea. It was the organizational structure deciding whether the idea got nurtured or strangled. Understanding those structures — and knowing how to build them for loonshot development — is one of the more important capabilities a leader can have. Bahcall hands over the framework. The rest is on whoever’s holding it.
A Note on the Book’s Own Form
That it succeeded — that it became a widely read, genuinely influential book — is itself evidence for the model. Ideas that challenge existing frameworks, that offer something actually new instead of repackaged wisdom, that demand more intellectual engagement than the average product in their category, can succeed when they’re developed with enough care and communicated with enough skill. Bahcall developed his with care and communicated it with skill. The result is a book that rewards re-reading as a career develops and organizational responsibilities shift, because the questions it raises — about incentive structures, phase transitions, the balance between exploration and exploitation — are perennial leadership problems that show up differently at every scale and in every context.
Whether you’re running a startup or a division of a giant corporation, managing a research team or running a school, the underlying question Bahcall raises stays the same: have the structural conditions been built that let your most talented people chase the most important unconventional ideas — or has a structure gotten built, unintentionally, that rewards franchise defense and punishes loonshot development? The answer determines more about long-term organizational success than any other single factor, and Loonshots hands over both the framework to answer honestly and the tools to change the answer if the current one isn’t good enough.
Final Thoughts
Management books come and go, most of them recycling the same advice in new packaging. Loonshots is different because it offers something genuinely new: a causal model pulled from physics that explains organizational behavior with more precision than most social science frameworks manage, tested against carefully chosen historical cases, and translated into specific, actionable structural prescriptions. Bahcall did something rare in popular management writing — built an actual theory instead of just collecting observations — and the theory is rigorous enough to be useful and accessible enough to actually get read.
The organizations that produce the next decade’s transformative innovations — the drugs that treat what’s currently untreatable, the technologies that crack problems currently uncrackable, the approaches to education and governance that let people flourish at scales currently out of reach — will be the ones that solved the loonshot problem. They’ll have found ways to hold the balance between exploration and exploitation, to shield early-stage unconventional ideas from premature franchise judgment, and to build the structural conditions that make chasing a loonshot a rational career move rather than an act of institutional martyrdom.
Bahcall wrote the most useful guide currently available for solving that problem. It’s earned the wide readership and influence it’s gotten. Read it with a pencil, ready to underline whatever diagnoses your own organization’s dynamics, and to note the structural moves that might actually improve them. The physics is elegant, the history is gripping, the prescriptions are specific. That’s everything a management book is supposed to be and mostly isn’t.
The loonshots that matter most are the ones that nearly didn’t happen. Understanding why they nearly didn’t, and what structural conditions would have made survival more likely, is one of the more important contributions any management thinker can make. Bahcall made that contribution with rigor, clarity, and real originality. The field is better for it.
For leaders who want their organizations to stay capable of the unconventional ideas that determine long-term survival, and who want something more rigorous than “embrace innovation culture,” Loonshots is where to start. Its mix of theoretical precision, historical depth, and practical prescription makes it one of the more useful books on organizational design published in the last decade, and its core insight about phase transitions and structural incentives will stay relevant as long as organizations exist and the tension between exploration and exploitation persists — which is to say, indefinitely.
The book earned its place as a widely-cited work in organizational behavior through the quality of its thinking, not marketing or celebrity authorship. That quality — a physicist’s rigor turned on a management problem, a historian’s command of evidence used to validate theory, a practitioner’s experience shaping the prescription — is rare enough in any field to be worth noting. Bahcall wrote a book that stands apart from the noise on the management shelf and rewards the sustained attention it asks for. In doing so he produced something that is itself, in the best sense, a loonshot: an unconventional contribution that challenges the existing framework and offers something genuinely new in its place.
Related: Churchill: Walking with Destiny Summary
The Practical Framework: Applying Loonshots Summary In Real Life
The Phase Transition Model: Why It Explains What Culture Explanations Miss
Every large organization that’s ever killed an important innovation has, at some point, diagnosed the problem as cultural. The language is always the same: we need to be more entrepreneurial, more willing to take risks, more comfortable with failure, more like a startup. These diagnoses aren’t exactly wrong — the cultural symptoms are real enough. But Bahcall’s contribution is showing that culture is the effect of structural incentives, not the cause of them. Trying to fix an organization’s innovation culture without touching the structural incentives that produced it is a bit like trying to change water’s temperature by changing how you feel about it.
The actual lever is somewhere else entirely.
The phase transition metaphor rewards unpacking, because its precision is where the practical value lives. In physics, a phase transition is a state change — water to ice, iron to magnet — that happens suddenly once a specific threshold gets crossed, not gradually. Below the transition temperature, water is liquid, molecules moving freely. Above it, they lock into the rigid structure of ice. The underlying physics hasn’t changed. Only the organizing behavior of the molecules has. And the transition is driven by a quantitative shift in one variable — temperature — that produces a qualitative shift in the whole system’s behavior.
Bahcall names the organizational equivalents: the equity fraction of an employee’s compensation, which he calls E, and the fitness landscape — his term, not one this rewrite is endorsing — of the organization, which he calls S, for “spread” or cultural noise. What matters practically: as organizations grow, the political payoff for managing rank and relationships rises relative to the payoff for working on loonshots. Not a failure of character. Arithmetic. A successful loonshot might add ten percent to company value. A successful political move might raise one executive’s power, pay, and status by far more, and far faster. Cross a certain threshold of size and structure, and the incentive math tips — and the organization’s collective behavior, not through any single decision but through the aggregated choices of rational people, tips right along with it, from loonshot-nurturing to franchise-protecting.
Which is what makes phase-transition thinking so much more useful than the culture explanation. Culture explanations imply a motivational fix: get people to care more, run a workshop on creative risk-taking, celebrate failure out loud. Phase-transition thinking implies a structural fix: shift the incentive terrain so the transition threshold itself moves. That means touching equity structures, span of control, promotion criteria, the organizational wall between exploratory and exploitative work. Engineering problems with engineering solutions — harder to pull off than a culture workshop, more likely to actually work.
The practical implication for anyone leading: ask the structural question before the motivational one. When the organization fails to protect an important idea, ask first — what structural incentives are working against it surviving? Who benefits from killing it, blocking it, deprioritizing it? What organizational separation would have bought it more time before getting judged by the franchise’s criteria? These questions produce answers you can act on. “How do we get people more comfortable with risk?” does not.
P-type and S-type Loonshots: Why the Distinction Changes Your Innovation Strategy
One of Bahcall’s subtler contributions is splitting loonshots into two fundamentally different types — P-type and S-type. Most innovation writing treats all innovations as roughly the same: new ideas facing organizational resistance, needing champions and air cover to survive. Bahcall’s split shows that different types of innovation fail for different reasons, need different development processes, and need different kinds of organizational protection.
P-type loonshots are product innovations — new technologies, new compounds, new devices. The transistor was a P-type loonshot. Penicillin was a P-type loonshot. The iPhone was a P-type loonshot. What distinguishes them: the innovation is primarily technical. A capability that didn’t exist before now does, and its value hinges mostly on whether that technical capability is real and can be made reliable and manufacturable at a price that makes sense.
S-type loonshots are strategy innovations — new ways of organizing, new business models, new routes to market, new ways of reading the competitive field. Walmart’s hub-and-spoke logistics was an S-type loonshot. The fast-food franchise model was an S-type loonshot. Direct-to-consumer models in industries where middlemen had always run distribution are S-type loonshots. What distinguishes them: the innovation is organizational or strategic, a new way of doing something that already exists, and its value hinges on whether the new approach actually beats the old one in practice.
The split matters because S-type loonshots are much harder to protect from being judged too early. A P-type loonshot can survive organizational skepticism as long as it survives in a lab — the technical reality of a new capability is a fact that skeptics eventually have to swallow whether they like it or not. An S-type loonshot lives and dies in implementation, which leaves it exposed to franchise judgment at every stage. The people defending the existing strategy always have the same argument on hand: we know the old way works. We don’t know the new way works. Every scrap of early evidence that the new approach isn’t working yet — which is almost always true of new strategies early on — becomes ammunition to kill it before it’s had time to be fairly judged.
Different implications for each. P-type loonshots need technical air cover — shelter from the commercial side’s demand for immediate practical payoff. S-type loonshots need sponsorship at the very top, because only leaders with the authority to override the franchise’s legitimate objections can give an S-type loonshot the implementation runway it needs to develop. Organizations that get this distinction can build more precisely targeted protection for whichever type of innovation they’re actually trying to grow.
The Moses-Bush Roles: How Great Innovators Actually Operate
Bahcall’s read on the innovator’s role is one of the more counterintuitive parts of the book, and it cuts directly against the dominant mythology of innovation leadership. That mythology says: the great innovator is the visionary who champions an idea against organizational resistance, refuses to hear no, keeps pushing until the organization finally sees the value. It’s the Steve Jobs story, the Elon Musk story, the Thomas Edison story — lone genius versus the conservative establishment.
The biographical record Bahcall works through suggests this is mostly wrong, or at least badly incomplete. The people who successfully carried important ideas through large organizations to world-changing impact were usually playing two roles at once — and the second one barely gets mentioned in the mythology at all. Bahcall calls these the Moses role and the Bush role, borrowing the biblical imagery of Moses leading his people out of Egypt and the burning bush that spoke to him without being consumed by its own fire.
The Moses role is the one the mythology loves: the leader who rallies people around the vision and drives the charge toward the promised land. It takes charisma, conviction, urgency, the ability to get people to follow into territory nobody’s mapped yet. Without it, important innovations never build the organizational momentum to actually get implemented at scale.
The Bush role is quieter, and rarer: staying analytically honest about the current state of the evidence, updating on new information, distinguishing what’s known from what’s merely hoped, giving the people working on the loonshot real feedback instead of blind encouragement. The Bush leader asks the hard question. The Bush leader is the one who can tell a team their loonshot isn’t working yet without gutting their motivation, and tell them exactly what needs to change to improve its odds.
The leaders who actually drove important innovations through to impact — Bahcall examines Vannevar Bush himself, along with several corporate and military figures — were the ones who could play both roles, switching as the moment demanded. They inspired and championed when that’s what was needed. They gave honest analytical feedback when that’s what was needed. And, crucially, they could tell which role the moment called for. Organizations stocked only with Moses types get inspiring champions who lose the thread the moment the evidence complicates the story. Organizations stocked only with Bush types get honest analysts who can’t generate the momentum real change requires. The combination is what produces world-changing results, and it’s rare enough to explain why world-changing innovations stay rare too.
Applying Loonshots to Your Own Organization
The gap between Bahcall’s framework and the actual innovation problems most organizations face is smaller than it looks. Nobody needs to be running a pharmaceutical company or a military research program to apply this. The basic dynamics — phase transitions driven by incentive structure, the tension between loonshot and franchise, the need for separation and dynamic equilibrium — operate at every scale an organization can take.
At the team level: find out who’s working on the equivalent of a loonshot — new approaches, new methods, ideas that challenge current practice — and take an honest look at the structural pressure they’re under. Are they judged on the same metrics as the people maintaining the current approach? If so, the math is almost always stacked against their loonshot surviving long enough to be tested fairly. Even a partial separation — a different evaluation timeline, different success criteria, a different review process — changes the incentive terrain enough to save more of the promising ideas.
At the organizational level: pull up the last three to five important innovation failures. Ask Bahcall’s question for each one — did the idea fail because it was actually bad, or because the organization’s structural incentives worked against it at exactly the moment it was most vulnerable? Be honest about it. The franchise almost always beats the loonshot when they compete on equal footing, because the franchise has proven value and the loonshot only has potential value. Equal footing is a rigged game for loonshots, every time.
At the strategic level: name one P-type and one S-type loonshot the organization is currently incubating, or should be, and map out the separation mechanisms protecting — or that should be protecting — each. For the P-type: who’s providing technical air cover against commercial pressure for immediate payoff? For the S-type: who at the executive level is providing the sponsorship that lets it develop against the franchise’s legitimate objections? If the honest answer to either is “nobody, specifically,” the loonshot’s survival comes down to luck rather than design. Design is the more reliable bet, and Bahcall has handed over the design principles.
References
Editorial StandardsCorrectionsMedical DisclaimerAbout Our ContentAffiliate DisclosureSite Map