Relevant For:
- Managers who must choose under uncertainty and want a repeatable approach
- Analysts tasked with turning soft data into numerical inputs
- Risk officers and product leaders deciding where to invest measurement effort
Remarkable Points:
- Frames measurement around value of information: measure what changes the decision, not what is interesting.
- Practical methods to quantify so-called intangibles using sampling, calibration, and Bayesian updates.
- Shows that small, cheap measurements often yield outsized reductions in uncertainty.
- Introduces simple decision mathematics that connect measurement effort to economic payoff.
- Emphasizes clear operational definitions to avoid ambiguous metrics and wasted effort.
Most Liked (As-Is Statements):
- "If it affects a decision it is measurable in a way that matters."
- "Calibrated judgment beats confident guesswork."
- "Small experiments change probabilities, and probabilities change choices."
Key Themes:
Theme 1: Value of Information
Measurement is valuable only to the extent it alters decisions. The book forces the reader to tie every proposed measurement to expected economic benefit, pruning trivial metrics.
Theme 2: Operational Definitions
Ambiguity is the real blocker. Hubbard insists on precise definitions so measurements are reproducible and directly relevant to the question at hand.
Theme 3: Practical Uncertainty Quantification
Techniques range from calibrated expert elicitation to Monte Carlo simulation. The emphasis is on making uncertainty explicit and actionable, not on mathematical elegance.
Theme 4: Cost-effective Measurement
There is a calculable point of diminishing returns. The book teaches how to estimate the marginal value of information and stop when additional measurement is not justified.
Final Remark:
How to Measure Anything is a toolkit for decision-centered measurement. It pares away ritual metrics and replaces them with purposeful, economic thinking. The value lies in its mental models: define the decision, quantify what matters, and spend only to the point where uncertainty still matters. Read it for the discipline it imposes, not for entertainment.
Measure to change choices, not to impress with numbers.
Similar Books:
- The Signal and the Noise by Nate Silver
- Superforecasting by Philip E. Tetlock and Dan M. Gardner
- Thinking in Bets by Annie Duke
- How Not to Be Wrong by Jordan Ellenberg