Small-Bet Tail Strategy
Take many survivable swings so rare winners can dominate results
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 5
- Confidence
- 95%
Housel describes tail outcomes: a small number of events can explain most of the result. In venture portfolios, public markets, and innovative companies, many attempts may fail or merely perform adequately while one or two winners pay for the rest. Because those winners are hard to identify in advance, the method is not to bet everything on a confident forecast. Instead, create multiple bounded attempts, keep each loss survivable, and preserve the ability to continue. A nonzero failure rate becomes evidence that the portfolio is exploring rather than repeating only safe work. When a winner appears, resources can concentrate around it. The discipline lies in sizing, not celebrating failure for its own sake. Failed experiments should produce information or preserve optionality, and the portfolio must never require every attempt to succeed.
Origin
Extracted from The Diary of a CEO
Core principles
- 01A few outcomes can drive most total returns
- 02Future winners are difficult to identify beforehand
- 03Failure is expected evidence, not automatic incompetence
- 04No single experiment should threaten the ability to continue
How to run it
- 1
Define the tail opportunity
Choose a field where a rare success could materially outweigh several small failures.
Watch out Do not assume every domain has venture-style payoff distribution.
- 2
Cap each loss
Set time, money, and reputation limits that keep one failed attempt from ending the portfolio.
Pro tip Make the stop condition explicit before enthusiasm grows.
- 3
Generate enough variation
Run multiple distinct attempts rather than repeatedly packaging the same idea.
Watch out Volume without meaningful variation does not create useful optionality.
- 4
Accept informative failure
Use misses to update assumptions and verify that the portfolio is taking some real risk.
Pro tip Record what the failure changed about the next bet.
Watch out Failure is not valuable when it violates a hard boundary or teaches nothing.
- 5
Back the emerging winner
When evidence identifies an outsized result, shift more attention and capital toward it.
Watch out Do not confuse one noisy early result with a durable winner.
In the wild
Bartlett and Housel discuss Amazon's failed Fire Phone alongside successful bets such as AWS. Housel's point is that a company willing to run many bounded experiments can tolerate visible misses because a small number of very large wins may dominate the portfolio.
→ The portfolio's success depends on rare large outcomes rather than a perfect hit rate.
Housel recounts a report claiming that all of Netflix's new movies had succeeded. The CEO reportedly treated that as bad news because a perfect success rate suggested the company was not taking enough creative risk.
→ A controlled failure rate served as a signal that experimentation remained real.
Common mistakes
Making one existential bet
A tail strategy requires repeated attempts; one ruinous wager removes the mechanism.
Rewarding failure without learning
A miss only helps when it improves the portfolio or preserves a credible next attempt.
Demanding a perfect hit rate
Avoiding every visible failure can also avoid the novelty required for an outsized success.
Is it for you?
Best for
It is best for product, content, or investment portfolios where outcomes are highly skewed and experiments can be bounded.
Not ideal for
It is not ideal where each failure carries irreversible safety, legal, ethical, or financial consequences.
From the transcript
“just a couple of things that happen explain 90 or 99% of what matters”
“make sure your bets are not that big and you can just keep taking little risks”
From the episode
The Savings Expert: “Do Not Buy A House!”, How To Turn £100 Into £1.5m Without Effort: Morgan Housel