Variability as an Asset

Definition

Reinertsen distinguishes the amount of variability (a statistical property) from its economic cost (which depends on the payoff-function it passes through). Where lean manufacturing correctly treats variability as waste — because repeating an identical product creates value every time — product development creates value only by changing the design, which necessarily introduces uncertain outcomes. "We cannot add value without adding variability, but we can add variability without adding value." The chapter's central claim: variability should be neither minimized nor maximized, but exploited when the payoff function is asymmetric.

In the Book

Chapter 4's opening example offers three technical investigation paths, each costing $15,000: Path A (50% chance of a $100,000 payoff), Path B (90% chance of $20,000), Path C (100% chance of $16,000). Path C has zero uncertainty but the lowest expected value ($1,000); Path A has the most uncertainty and the highest expected value ($35,000) — proof that minimizing variability and maximizing economic outcome are different goals. Principle V2 imports the Black-Scholes option-pricing logic: an option's asymmetric payoff (unlimited upside, capped downside at the premium paid) means that increasing volatility increases the option's value, the reverse of the manufacturing intuition. Reinertsen applies this directly to a pharmaceutical company choosing between a well-understood 20-year-old molecule and a poorly understood one freshly extracted from Amazon mud: since failure only costs a $10,000 experiment while success can yield a billion-dollar drug, the high-variability candidate is the economically superior choice, not the riskier one to be avoided.

Why It Matters

This concept undercuts the reflexive transplant of manufacturing quality doctrine (zero defects, Six Sigma, minimize variance) onto any activity that is actually about generating new information rather than repeating a known process. It gives a precise test for when variability should be embraced: does the activity have an asymmetric payoff, where the downside is capped and the upside is not? If so, the lever to pull isn't "reduce risk" but "reshape the payoff function" — cap the downside (fast feedback, small experiments) while leaving the upside open — which changes how R&D portfolios, hiring bets, and any exploratory investment should be evaluated.