Skip to main content

Power Law

Are you concentrating effort where returns are asymmetric?

The biggest secret in venture capital is that the best investment in a successful fund equals or outperforms the entire rest of the fund combined — Peter Thiel

Perceive

The power law says a small number of inputs produce the majority of outputs. Not 50/50. Not even 80/20. More like 90/10 or 99/1 in the domains that matter most.

This is not a suggestion to prioritize. It's a mathematical fact about how returns distribute in complex systems. One customer segment drives most revenue. One channel produces most leads. One product feature retains most users. One decision determines most of the outcome.

Normal vs Power Law

Normal DistributionPower Law
ShapeBell curve — most outcomes cluster around the meanLong tail — a few outcomes dwarf the rest
StrategyDiversify, spread risk evenlyConcentrate, find the outlier
ErrorOverfitting to extremesAveraging away the signal
ExampleHeight of adultsRevenue per customer

Where Power Laws Rule

DomainThe FewThe Many
Venture capital1 investment returns the fund9 out of 10 fail
Sales pipeline1 channel produces 80% of leads5 channels split the remaining 20%
Product features1 feature drives retention20 features are used once
Content1 article drives 50% of traffic100 articles split the rest
CustomersTop 10% produce 50%+ of revenueBottom 50% barely cover costs
Employees1 hire changes the trajectoryMost hires maintain it

Question

Why do people allocate resources as if returns were normally distributed?

Because equal allocation feels fair. Spreading effort across 10 things feels safer than betting on 1. But power law returns punish diversification. The cost of missing the outlier exceeds the cost of all other mistakes combined.

The Allocation Trap

StrategyFeels LikeActually Is
Spread evenly across 6 venturesPrudent diversificationGuaranteed mediocrity
Focus on the 1 that shows signalDangerous concentrationPlaying the math
Give every channel equal budgetFair resource allocationStarving the winner
Treat all customers equallyGood serviceSubsidizing the unprofitable

Thiel's Question

"What important truth do very few people agree with you on?"

The power law applied to beliefs: one contrarian insight, if correct, is worth more than a thousand consensus opinions. The returns to original thinking are asymmetric.

Act

Find the Outlier

StepActionSignal
1. Measure everythingTrack returns per input across all activitiesData, not intuition
2. Rank by outputSort channels, customers, features by actual returnTop 3 vs bottom 50
3. Double down on #1Shift resources from average to exceptional2x investment in the winner
4. Cut the bottomStop subsidizing what doesn't workFree up capacity
5. Look for the next outlierThe power law applies to the search itselfMost searches fail, one transforms

Power Law Audit

AreaWhat's your #1?% of total resultGetting enough investment?
Revenue source
Customer segment
Marketing channel
Product feature
Team member
Time allocation

Decision Rule

When allocating resources across competing options:

  1. If one option shows 3x+ better returns than others, stop splitting evenly
  2. If you can't identify a clear winner, run small experiments until one emerges
  3. If everything performs equally, you're probably measuring the wrong thing

Checklist

  • Do you know which 20% of inputs produce 80% of your results?
  • Are you investing proportionally in your top performers?
  • Are you subsidizing underperformers out of fairness instead of math?
  • Have you killed your worst-performing activity in the last 90 days?
  • Are you searching for the next outlier or maintaining the average?

Context

Questions

If you could only keep one activity in your business and kill everything else, which one would you keep — and why aren't you spending most of your time on it?

  • What would your allocation look like if you took the power law seriously instead of hedging?
  • Which of your six ventures shows the strongest signal — and is it getting proportional investment?
  • Are you diversifying because the math supports it, or because concentration feels risky?