Bullets Then Cannonballs
Validate with small bets before committing big resources
Bullets Then Cannonballs is Collins's framework for disciplined risk-taking based on his research into companies that thrived in uncertainty. A bullet is a low-cost, low-risk, low-distraction test that helps calibrate your aim. You fire multiple bullets to figure out what works, gathering empirical evidence before committing major resources. A cannonball is a concentrated bet of significant resources behind a validated concept. The key discipline is firing bullets first to gain empirical validation, then concentrating resources into cannonballs once you have evidence. Companies that fired uncalibrated cannonballs — making big bets without empirical validation — often destroyed themselves. Companies that only fired bullets but never concentrated into cannonballs missed their windows of opportunity. The magic is in the sequence: bullets first, calibrate, then cannonballs.
- Small empirical tests reduce risk more effectively than extensive planning or analysis
- Validation must come from real-world evidence, not projections or market research
- The sequence matters: bullets first, then cannonballs
- Firing only bullets without ever concentrating into cannonballs is just as dangerous as firing uncalibrated cannonballs
- Design Low-Cost BulletsBefore committing significant resources to any initiative, design small experiments that can test your core assumptions with minimal investment. A bullet should cost less than 5% of your total available resources and take less than 10% of your available time. The goal is not to build a scaled version — it is to generate empirical evidence about whether your direction is correct. List three to five small experiments you can run within the next two weeks.Pro tipThe best bullets test your riskiest assumption first, not your easiest oneWarningDo not gold-plate your bullets — the point is speed and learning, not perfection
- Calibrate Based on EvidenceAfter firing bullets, objectively analyze the results. Which bullets hit something promising? Which missed entirely? Calibration means honestly assessing what the data tells you, not what you hoped it would say. Look for bullets that generated unexpected positive results or strong customer signals. Be willing to abandon directions where bullets consistently miss, regardless of how much you believed in the idea beforehand.Pro tipKeep a written log of every bullet fired and its specific results to prevent revisionist historyWarningConfirmation bias is your biggest enemy during calibration — seek disconfirming evidence actively
- Fire Concentrated CannonballsOnce you have empirical evidence from successful bullets, concentrate significant resources into a cannonball aimed at the validated target. This is where you go big — hiring, investing, building infrastructure, committing fully. The discipline is that your cannonball must be aimed at a target that has already been validated by bullets. Do not fire cannonballs based on hope, theory, or someone else's success. Your own empirical evidence must justify the concentration of resources.Pro tipEven with validated targets, maintain enough reserve resources to fire additional bullets if the cannonball needs mid-course correctionWarningThe most common failure mode is executives who get excited by an unvalidated idea and fire a cannonball without ever testing with bullets
Apple fired the iPod as a calibrated bullet into the music market, testing whether consumers would adopt Apple's approach to digital media. Once validated, they fired the iTunes Store cannonball, then used that validation to fire the iPhone cannonball into mobile computing. Each major bet was preceded by empirical validation from the previous initiative.
Collins developed this concept in Great by Choice while studying companies that thrived in chaotic, uncertain environments. He used the metaphor of a ship battle: imagine you have limited gunpowder. You could fire a massive cannonball but if you miss, you have wasted your resources. Or you could fire small bullets to calibrate your aim, and once you know the trajectory is right, fire a concentrated cannonball. The research revealed that the 10X companies consistently followed this discipline while their comparison companies often fired uncalibrated cannonballs that missed their targets catastrophically.