How to Measure Anything

This book is not “clickbait”. The author does, indeed, explain how to measure anything.

But even more importantly, the author asks the important questions. What is measurement? Why would you measure?

The questions are related, and the second is the one that should be answered first. I bought this book as part of buying a few books on practical decision theory. This book came up in the search.

The reason to measure is to make better decisions. If decisions are not informed by the measurement, then there is no need to measure. This means that the only relevant meaning for “measurement” is in the context of a decision process.

The author’s definition for measurement is “a process that reduces your uncertainty about something”. The author repeatedly explains that measurement is inherently a process with errors. If you have no idea how big a door is, and you compare it to your arm, you have reduced uncertainty.

You will reduce uncertainty even more if you pull out a tape measure. This takes more effort, but still involves some error. You can reduce it even more with sophisticated laser-based equipment. This takes even more effort, and even more effort if you want to use a blue-light laser, which has even tighter uncertainty.

So what should you do?

The answer is simple. It depends on why you need to know the door’s size. If you need to move a couch in, and you need to know if it fits, using your arms is probably fine.

If you need to cut some wood to replace the frame, a tape measure is probably the tool: the uncertainty using arms is too much.

This ties closely to the concept of “value of information”, which the author also explains.

But beyond the theory, and even beyond the practical advice, of which there is a lot, there is one theme throughout the book. You can measure things. If they matter, you probably should. And you probably can.

From the monetary loss of brand value because a spokesperson tweets something bad to the amount to spend saving a life, you can, and should, measure it. Almost always, the people who say that you cannot put a monetary value on human life undervalue it compared to those who measure it.

Measuring is a moral imperative. Getting better at it is ethically important.

I do feel sorry for the author for having to explain how to program using Excel formulae, because business people reject anything more sophisticated.