The Out-Fail the Competition System
A systematic approach to increasing your rate of experimentation and failure, based on the
A systematic approach to increasing your rate of experimentation and failure, based on the principle that failure rate correlates directly with success rate. Companies like Amazon, Booking.com, and Tesla succeed by running orders of magnitude more experiments than their competitors. The framework includes five organizational principles: remove bureaucracy, fix incentives, promote fast-failers, measure experiment rate, and share learnings from failures.
- The rate of experimentation is a stronger predictor of long-run success than the quality of any single bet.
- Organizations that punish failure starve themselves of the feedback loop that produces eventual success.
- Bureaucracy compounds the cost of each experiment and should be eliminated before attempting to increase experiment volume.
- Sharing failure learnings openly turns each loss into a distributed asset rather than a buried liability.
- The goal is not to fail less but to fail faster, cheaper, and more informatively than competitors.
- Classify decisions as Type 1 or Type 2Use Jeff Bezos's framework: Type 1 decisions are irreversible one-way doors (proceed carefully). Type 2 decisions are reversible two-way doors (proceed quickly). Most decisions are Type 2 but get treated as Type 1, causing organizational paralysis. Identify which type each decision is before determining speed.Pro tipGet to 51% certainty and make the decision. As Obama said about deciding to pursue Bin Laden: 'You don't have to get to 100% certainty on your big decisions, get to 51%, and when you get there, make the decision quickly.'
- Remove bureaucracy from Type 2 decisionsMake project teams as small as possible. Give them authority, trust, and resources. Cut back sign-off processes for reversible decisions. Haier Group divided 75,000 employees into 4,000 micro-organizations of 10-15 people each, enabling lightning-fast decisions.WarningThe real cost of indecision is not avoiding failure - it is wasted time, missed opportunity, and lost learning. By the time one company spent 10 months deciding, the opportunity had disappeared.
- Fix incentives to reward experimentation, not just outcomesRecognize and reward employees when experiments are successfully EXECUTED, regardless of outcome. Executing the experiment is the controllable factor; market success is not. Writing 'fail faster' on the kitchen wall means nothing if people are still incentivized to just do their job and avoid risk.Pro tipAsk your team: 'What reason do you have to fail more often?' Then ask: 'What reason do you have NOT to fail?' The second list reveals your misaligned incentives.
- Promote fast-failers and remove blockersIdentify employees with the highest experiment rates and promote them. Every manager creates a sub-culture - you need fast-failers at the top to cascade the philosophy downward. Conversely, swiftly remove individuals who obstruct experimentation, especially managers.Pro tipInfluence trickles down: with 30 managers, you really have 30 company cultures. One bad manager can destroy the entrepreneurial spirit of an entire team.
- Measure and share failure rate as a KPIEstablish a clear experimentation process everyone follows. Measure each team's failure rate with a goal of increasing it by a factor of 10. Share the details of every failed experiment across the organization to prevent duplication and spark new ideas. Booking.com runs 1,000 experiments at any given moment.Pro tipExpect roughly 3/10 experiments to be tragic failures, 3/10 average failures, 3/10 good, and 1/10 great - but the one great result pays for all the other nine.
Treating all decisions like Type 1 (irreversible)
As organizations grow, they tend to apply heavyweight decision-making to all decisions, including easily reversible ones. This causes slowness, risk aversion, insufficient experimentation, and diminished invention. Most decisions are two-way doors.
Writing values on walls instead of encoding them in incentives
Putting 'fail faster' in the employee handbook is meaningless if people are incentivized to avoid embarrassment, play it safe, and just do their job. Human behavior is driven by incentives, not platitudes.
This framework comes from Law 21: You Must Out-Fail the Competition in Steven Bartlett's Diary of a CEO.
Source · BOOK
The Diary of a CEO