To measure anything, including things often considered "intangibles," you follow a process of reducing uncertainty through observation and quantification. Key points from Douglas Hubbard's book How to Measure Anything explain this approach:
- Anything observable can be measured: If you care about something, it must be detectable in some way, which means it can be observed in some amount and thus measured. Measurement is defined as a quantitatively expressed reduction of uncertainty based on one or more observations
- Measurement does not require perfect precision: Even a fuzzy or small-sample measurement that reduces uncertainty is valuable. For example, small sample sizes can still provide meaningful insight, such as the "Rule of Five," which states a 93.75% chance that the median of a population lies between the smallest and largest values in a sample of five
- Start with a clear problem definition: A well-stated problem is half solved. Define what decision the measurement is meant to support and what specifically needs to be measured and to what degree of accuracy
- Use simple observations and clever calculations: Historical examples like Eratosthenes measuring Earth's circumference with basic observations and Fermi's intuitive estimation techniques show that complex tools are not always necessary
- Apply practical methods to intangibles: The book provides methods to measure things like customer satisfaction, risk, technology ROI, and organizational flexibility, which are often seen as immeasurable
- Leverage Bayesian reasoning and information economics: These help in simplifying measurement problems and making the most of limited data by focusing on reducing uncertainty relevant to decisions
In summary, measuring anything involves recognizing that all important variables can be observed in some form, defining the measurement problem clearly, using even small or indirect data to reduce uncertainty, and applying practical, often simple, methods to quantify intangibles and complex phenomena
