The Magic Number (Innovation Equation)
A mathematical model for when organizations flip from innovation to politics
The Magic Number is Bahcall's mathematical model explaining why organizations undergo a phase transition from nurturing innovation to killing it as they grow. Just as water freezes at 32 degrees Fahrenheit because the binding forces between molecules overcome the entropy that keeps them moving, organizations flip from innovation to politics at a critical size because the career incentives for political maneuvering overcome the project incentives for doing good work.
The equation shows that this critical size M (the magic number) depends on four control parameters: equity fraction (E) -- how much of compensation comes from equity stakes in the organization's success rather than salary; management span (S) -- the number of people reporting to each manager; organizational fitness (F) -- the ratio of return on skill to return on politics, measuring how much project quality versus political lobbying determines promotions; and salary step-up (G) -- how much compensation increases with each promotion. For typical real-world values, the magic number falls around 150, explaining the empirical observation (from Dunbar's research and historical examples like the Mormon exodus) that groups transform at roughly this size.
Critically, unlike Dunbar's biological hypothesis about brain-size constraints, the Innovation Equation reveals concrete levers that leaders can adjust to raise the magic number. By increasing equity fraction, widening management spans in innovation groups, improving organizational fitness (making promotions more merit-based and less political), and flattening salary curves, organizations can keep larger groups in the innovative phase longer.
- Organizations undergo a phase transition at a critical size, just as water freezes at a critical temperature: it is a structural inevitability, not a cultural failure.
- The magic number at which organizations flip from innovation to politics depends on four adjustable control parameters, not on any fixed biological or cultural limit.
- Reducing return-on-politics -- making it harder for political maneuvering to influence compensation and promotion -- is the single most powerful lever for raising the magic number.
- The equity fraction of compensation aligns individual incentives with organizational success; increasing it raises the magic number.
- Middle managers are the weakest point in the battle between innovation and politics because they are close enough to the top to benefit from promotion but disconnected enough from projects to be tempted by political gamesmanship.
- Audit your current control parametersAssess your organization's current state on each of the four control parameters: What fraction of total compensation comes from equity or outcome-based components (E)? What is the typical management span (S)? How much does compensation increase with each promotion (G)? How much do politics versus skills determine promotion decisions (F)?Pro tipThe most revealing diagnostic is to ask employees at the middle-management level: what percentage of your time do you spend on project work versus managing up and navigating organizational politics? If the answer skews toward politics, your magic number is already too low.
- Reduce return-on-politicsMake lobbying for compensation and promotion decisions more difficult. Find ways to make those decisions less dependent on an employee's direct manager and more independently assessed and fairly calibrated across the company. Use peer review, cross-functional evaluation committees, and standardized criteria.Pro tipBill Coughran at Google asked 180 engineers to report directly to him, deliberately creating such a wide span that he could not possibly micromanage or play favorites. The result was that engineers focused on their projects rather than managing their relationship with their boss.WarningReducing return-on-politics is culturally painful because it requires taking power away from managers. Many will resist, especially those whose authority depends on being the sole arbiter of their reports' careers.
- Increase the equity fraction and use soft equityShift compensation toward outcome-based components (stock, options, profit-sharing) that align individual incentives with organizational success. Also identify and apply nonfinancial rewards (soft equity) that strongly motivate employees: peer recognition, intrinsic motivators, meaningful project assignments, and public celebration of results.Pro tipCelebrate results, not rank. When the organization rewards the quality of project outcomes rather than the level of the person's title, it shifts the incentive balance toward innovation and away from politics.
- Increase project-skill fitInvest in the people and processes that scan for mismatches between employees' skills and their assigned projects. Help managers adjust roles or facilitate transfers between groups. The goal is to have employees stretched neither too much nor too little by their roles, maximizing the return on their technical skill.Pro tipWhen employees are well-matched to their projects, the return on devoting time to project work increases relative to the return on politics. This organically shifts behavior toward innovation.
- Fix the middle and fine-tune the spansIdentify and fix perverse incentives at the middle-management level, where the battle between innovation and politics is fiercest. Shift incentives from rewarding battles for promotion toward rewarding project outcomes. Additionally, widen management spans in loonshot groups (to encourage looser controls and more experimentation) while keeping narrower spans in franchise groups (to maintain operational discipline).Pro tipConsider hiring a chief incentives officer -- a specialist in incentive design -- to systematically audit and optimize the structural parameters that determine whether your organization rewards innovation or politics.WarningPerverse incentives at the middle-management level are often the unintended consequences of well-intentioned reward systems. They require careful diagnosis, not just blanket policy changes.
- Monitor and iterateTreat your organizational parameters as ongoing experiments, not one-time fixes. Regularly reassess the four control parameters, survey employees about the balance of project work versus politics, and adjust as the organization grows and market conditions change.Pro tipAs an organization grows, it must continuously adjust its parameters to stay ahead of the phase transition. What worked at 200 people will not work at 2,000.WarningThe phase transition is not a one-time event but a continuous threat. Even organizations that have successfully raised their magic number can drift back into the political phase if they stop actively managing their structural parameters.
The maker of Gore-Tex famously built new factories whenever a facility exceeded about 150 people, based on the intuition that innovation suffered in larger groups. They also used a flat organizational structure with wide spans of control, peer-based evaluation rather than top-down management, and significant profit-sharing.
Bill Coughran, a senior VP at Google, had 180 engineers reporting directly to him. With such a wide management span, he could not possibly micromanage or play favorites in promotion decisions. Compensation was assessed independently across the company rather than being primarily determined by a single manager's judgment.
DARPA (the Defense Advanced Research Projects Agency) maintained a structure that systematically raised the magic number: program managers serve short rotating terms (reducing political entrenchment), evaluation is project-based rather than rank-based, and the organization stays deliberately small while funding external innovation teams.
Bahcall, a physicist by training who ran a biotech company, recognized that the familiar observation of organizations losing their innovative edge as they grew was not just a cultural phenomenon but a mathematical one. He noticed that the patterns matched phase transitions in physics: not gradual decline but sudden snaps from one state to another.
He built a simplified mathematical model treating an organization as a tree of employees making rational choices between investing their time in project work (which increases the value of the organization's equity) and political maneuvering (which increases their likelihood of promotion and salary gains). The model produced an equation that predicted a critical organizational size -- a magic number -- at which the balance of these incentives would flip. This connected the empirical observations of Robin Dunbar (who found group-size limits around 150 through primate brain studies), the historical experience of Brigham Young (who organized Mormon companies of about 150), and the practical experience of W.L. Gore & Associates (which built new factories whenever a facility exceeded about 150 people).