Most strategic business errors stem from a single failure mode: importing a variance preference from one domain to another without noticing that the aggregation function changed. Sam Tomlinson's framework proves that uncertainty cannot be optimized globally. You must identify how outcomes are aggregated to choose the correct variance preference.
Multiplicative Compounds
Volatility Drag Penalty: In compounding operations (like COO or brand equity), variance compounds destructively. A 50% loss followed by a 50% gain leaves you at 75% (not 100%). Lower volatility at a moderate mean dramatically outperforms high averages with high volatility. Variance is punished.
Max-Order Extractions
Only the single best breakthrough peak performance counts; other results are discarded. In creative ad testing, product R&D, or venture portfolios, variance is your greatest asset. High volatility expands the right-hand tail, maximizing the winner you scale. Variance is rewarded.
Additive Sums
Outcomes accumulate linearly (like steady-state lead generation or standard channels). Over time, random fluctuations wash out via the Law of Large Numbers. High-variance swings add platforms learning friction without benefits. Focus strictly on arithmetic mean. Variance is neutral.
Hiring Lab: Volatility vs Role Optimization
Drag candidates to appropriate roles to optimize corporate output. Specialists are Drafts (Max-Order): peak ability counts; floors are ignored. Generalists are Chain Links (Multiplicative): consistency counts; low floors trigger bankruptcy. Optimizers are sums (Additive): only the average counts.
Available Talent Pool (Adjust sliders below!)
Brilliant viral marketing expert. Chaotic organizer, skips team meetings.
Solid communicator, exceptional planner. Lacks raw creative genius.
Math savant. Discovers ad bidding discrepancies. Poor listener.
Steady planner, consistent execution. Avoids controversial risks.
Brilliant inventor. Discovers extreme product breakthroughs, but completely ignores documentation.
Obsessive precision auditor. Prevents security exploits and product defects completely.
Volume content writer. Maintains steady output across channels with organic SEO rankings.
Highly empathetic support lead. Keeps customer churn extremely low with stellar communications.
Placement Zones (Drag Candidates Here)
Chief of Operations
Operations. Consistency matters. Volatility triggers system crash.
Creative Director
Ad testing. Requires extreme outlier swing. Floor is ignored.
PPC Manager
Always-on ads. Arithmetic mean matters. Volatility is neutral.
QA & Compliance Director
Security/Auditing. Weakest Link. Capped by worst-case Floor.
Multiplicative Compounding: Volatility Drag
Domains: Brand Trust, Long-Term Capital, Reputation.
A spectacular year followed by a catastrophic drawdown ruins you. In multiplicative systems, lower variance at a moderate mean dominates higher arithmetic averages.
Consistent Approach (Patagonia / Costco style)
High-Volatility Speculator (Constant Pivots / Hype style)
Volatility Drag Math
Notice the paradox: The speculative campaign has a much higher arithmetic average return (30% vs 15%), but because of its 80% volatility, its geometric compounding rate drags to **negative growth**, leaving it with less money than it started with!
Max-Order Extraction: Extreme Value Theory
Domains: Creative Ad Testing, Venture Portfolios, Specialist Hires.
You only scale the single best performer (the maximum) and discard the rest. In max-order aggregation, high variance is a massive asset because it stretches the right tail.
Extreme Value Formula
By taking wild, polarizing swings, the **Wild Pool has a lower average performance score (90 vs 100)**, but because of its higher variance, **its maximum peak (the winner you actually scale) dominates the safe pool!**
Additive Sums: Lead Generation Optimization
Domains: Always-on Lead Gen, Search Ops, Steady-State Activity.
Conversions sum up, and variance washes out over large sample sizes. Importing "max-order" draft logic here (whipsawing budgets weekly to chase last week's hero) destroys lead gen efficiency.
Arithmetic Summation
The whipsawed reallocator constantly shuts down steady channels because they didn't have a "hero" week, losing aggregate data. **The Steady-State Optimizer wins by focusing purely on arithmetic mean efficiency** and letting the Law of Large Numbers wash out weekly noise.