The Bush-Vail Rules
Four structural rules that let large organizations nurture radical innovation
The Bush-Vail Rules are four structural principles for sustaining radical innovation inside organizations, derived from the practices of Vannevar Bush (who organized US wartime science) and Theodore Vail (who built Bell Labs at AT&T). The core insight is that innovation failure is not primarily a culture problem but a structure problem: organizations undergo a phase transition as they grow, shifting from nurturing novel ideas to protecting careers and franchises. The solution is to design structure that manages this transition.
The four rules are: (1) Separate the phases -- create distinct groups for innovation ('artists') and execution ('soldiers'), each with tailored management tools. (2) Create dynamic equilibrium -- ensure a healthy, continuous exchange of projects and feedback between the two groups, with equal respect for both. (3) Spread a system mindset -- analyze decision-making processes, not just outcomes, to continuously improve how the organization makes choices. (4) Raise the magic number -- adjust incentive structures and organizational parameters so that politics remains subordinate to project work even as the organization grows.
These rules are not about cheerleading innovation or hiring visionaries. They are about engineering the conditions under which radical ideas can survive the organizational immune system that would otherwise kill them. Just as a physicist manages phase transitions by adjusting temperature and pressure, leaders manage innovation by adjusting structure and incentives.
- You cannot ask the same group to explore wild ideas and efficiently execute proven ones, just as you cannot ask water to be liquid and solid at the same time.
- Phase separation without dynamic equilibrium produces either chaos (no transfer of innovations to the field) or stagnation (no feedback from the field to the innovators).
- Analyzing decision-making processes matters more than analyzing outcomes because good outcomes can mask bad decisions and vice versa.
- The structural parameters of an organization determine its innovativeness far more reliably than motivational speeches or culture initiatives.
- Innovation is not about hiring the right visionary leader; it is about building the right structural conditions so that any good idea can survive.
- Separate your artists and soldiersCreate distinct groups for those working on radical, unproven ideas (artists/loonshot groups) and those working on proven, revenue-generating products and operations (soldiers/franchise groups). These groups need different physical spaces, different reporting lines, and fundamentally different management approaches.Pro tipTailor the management tools to each phase: wide management spans, loose controls, and flexible creative metrics for loonshot groups; narrow management spans, tight controls, and quantitative metrics for franchise groups.WarningSimply drawing a new box on the org chart and renting a new building is not enough. You must also tailor the incentive systems, review processes, and management styles to each group independently.
- Create dynamic equilibrium between the two groupsEstablish a natural, continuous process for projects to transfer from the loonshot nursery to the field, and for valuable feedback and market intelligence to cycle back from the field to the nursery. Both groups must feel equally valued. The leader's job is to manage the timing and quality of this transfer.Pro tipAppoint and train 'bilingual' project champions who can speak both artist-speak and soldier-speak, bridging the cultural divide between the two groups.WarningIf the transfer is overforced (a Moses commanding adoption), fragile early-stage ideas get permanently crushed. If underforced (no helping hand), promising innovations languish in the lab.
- Spread a system mindset across the organizationMove your teams from analyzing outcomes (why did this product fail?) to analyzing decision-making processes (why did we make the choices that led to this product?). Probe the people involved, data considered, analyses conducted, how choices were framed, and both financial and nonfinancial incentives at play.Pro tipThis process requires self-awareness, self-confidence to acknowledge mistakes, candor, and trust. It is often more efficient and less painful when mediated by a neutral expert from outside the team.WarningTeams naturally default to outcome mindset because it feels safer and more familiar. Without deliberate effort, system mindset will not take hold.
- Raise the magic number through structural adjustmentsUse organizational levers to push back the threshold at which politics overtakes project work. Key levers include reducing return-on-politics, using soft equity (nonfinancial rewards), increasing project-skill fit, fixing perverse incentives at the middle-manager level, and widening management spans in loonshot groups.Pro tipConsider engaging a chief incentives officer -- a specialist in the subtleties of incentive design -- to systematically identify and fix the structural parameters that determine whether your organization rewards innovation or politics.WarningPay special attention to the middle-manager levels, which are the weakest point in the battle between loonshots and politics. These managers face the strongest temptation to play politics because they are close enough to the top to benefit from promotion but far enough from the bottom to be disconnected from project work.
- Watch your blind side by nurturing both P-type and S-type loonshotsMake sure your loonshot nursery cultivates both product/technology innovations (P-type) and strategy/business-model innovations (S-type), especially the type your organization is least comfortable with. Most leaders and teams have a natural bias toward one type.Pro tipS-type loonshots are harder to spot because they involve small changes in strategy that no one thinks will amount to much. Deaths from S-type loonshots are gradual, making them especially dangerous.WarningIBM correctly anticipated the P-type loonshot of personal computers but missed the S-type loonshot of a software standard. Pan Am mastered P-type innovations in aviation but was blindsided by S-type operational efficiencies from competitors.
Bush created the Office of Scientific Research and Development, separating civilian scientists from military operators while maintaining close collaboration between the two. Scientists developed radar, the proximity fuse, and other critical technologies in their own loosely managed groups, while military operators provided feedback from the field and deployed proven innovations.
Vail restructured AT&T to include a separate research laboratory (eventually Bell Labs) that operated with different management principles from the telephone operating company. Researchers were given freedom to pursue basic science, while a systematic process transferred discoveries into commercial products.
Pixar created an environment of phase separation and dynamic equilibrium. Creative teams (artists) developed wild new story ideas while production teams (soldiers) executed proven filmmaking processes. The Braintrust -- a regular candid feedback session -- served as the mechanism for dynamic equilibrium, ensuring that early-stage ideas received honest feedback without being killed prematurely.
Vannevar Bush was an MIT engineer who persuaded President Roosevelt to create a new agency -- the Office of Scientific Research and Development -- to bridge the gap between military operators and civilian scientists during World War II. His structural innovations helped produce radar, the proximity fuse, and dozens of other technologies that turned the tide of the war. Theodore Vail, decades earlier, had built a similar structure at AT&T by creating Bell Labs as a separate research entity that maintained dynamic exchange with the operating telephone company. The research lab produced eight Nobel Prizes.
Bahcall synthesized their shared principles into four rules after recognizing that Bush and Vail had independently solved the same structural problem: how to keep radical innovation alive inside a large, operationally focused organization. Both men understood that innovation and execution require fundamentally different management approaches, and that trying to force both into the same structure guarantees the death of innovation.