How To Measure Anything

Book Review: Part I; How to Measure Anything by Douglas W. Hubbard

Dave Kinnear 1-On Leadership, Book Reviews

Image

Finding the value of intangibles in business:

Many times throughout my career, I’ve been confronted with the difficulty of presenting a business case for one project or another to colleagues or management. I suspect many of you have had similar experiences. Perhaps the most frequent reason for failing to win the approval of “my project,” was that I had not made a sound, measurable business case. Come to think of it, I have yet to have a project turned down by the management team or board when I properly demonstrated that it met ROI and strategic objectives. So the question is how do we make a sound, measurable business case for our projects? How do we measure risk with some of our more “soft” project components such as advertising and marketing results, new product initiatives, or, for example, simply evaluating whether to hire additional salespeople to grow market share? Enter Douglas Hubbard, author of How to Measure Anything. Hubbard has made what can be a deadly dull subject interesting and accessible. He writes in a clear and concise manner and uses many real-life examples to make his concepts “stick.” I wish I had access to the concepts in this book decades ago as I built my career in the semiconductor business.

I found several examples for measuring exactly what I needed in the past and always felt I could not measure. Specifically, I’m thinking about all the discussions I’ve had around the “risk” of adding a new salesperson to a territory – how much additional revenue will be generated? What about the market there? Where’s the economy going? What will happen if the new products aren’t introduced? What’s the financial risk of taking on the cost of adding that new salesperson? Getting my people to commit to a “number” so that I could assess the ROI was like pulling a hen’s teeth.  And to bring the point home, several of my colleagues asked me as recently as two months ago how to make this very decision. Luckily, I had just finished reading Hubbard’s book and was able to give them guidance; but more on this topic in Part II of this report. For now, let’s start with Hubbard’s definition of measurement:

Measurement: A set of observations that reduce uncertainty where the result is expressed as a quantity. As long as we are not willing to accept the best guess, or educated estimate, or range of possibilities for a difficult to measure item we will not move forward. Our decisions will be flawed. I found these four assumptions to be most useful:

  • Your problem is not as unique as you think
  • You have more data than you think
  • You need less data than you think
  • There is a useful measurement that is much simpler than you think.

Numbers can be used to confuse people; especially the gullible ones lacking basic skills with numbers. Therefore, we as leaders must be committed to making sure the whole organization is data-driven and understands the way we can reduce uncertainty through the straight forward techniques Hubbard explains. As he states, “The fact is that the preference for ignorance over even marginal reductions in ignorance is never the moral high ground.”

Hubbard gives us a very useful checklist for a Universal Approach to Measurement:

  • What are you trying to measure? What is the real meaning of the alleged “intangible?”
  • Why do you care — what’s the decision and where is the “threshold?”
  • How much do you know now — what ranges or probabilities represent your uncertainty about this?
  • What is the value of the information? What are the consequences of being wrong and the chance of being wrong, and what, if any, measurement effort would be justified?
  • Within a cost-justified by the information value, which observations would confirm or eliminate different possibilities? For each possible scenario, what is the simplest thing we should see if that scenario were true?
  • How do you conduct the measurement that accounts for various types of avoidable errors (again, where the cost is less than the value of the information)?

I especially enjoy the approach Hubbard takes to quantify the cost of making measurements based on the value of the information obtained. Too often, I have seen projects founder on either the inaction to get data which would be of great value and little cost or, perhaps, the exact opposite  — spending great amounts of time and money to obtain relatively useless information.

The key to putting the concepts in this text to work in your organization is to make sure that your team is “calibrated” for range estimates. I found that what Hubbard points out with respect to getting folks to “commit to a number” is true – they won’t if they can possibly avoid it. But getting them to sign up for a range of numbers is easier. Ranges such as sales will increase between $3 million and $5 million if we put a qualified new person in the field; the additional cost will be between $80 K and $100K for that person; gross margin should be between 45% and 50% are easier to think about. Yet, they present a problem since we are not at all sure how confident our team is in these numbers. Hence, calibration! Hubbard takes a great deal of time and care in providing the reader with an explanation of why we need to calibrate our team, what the goal is (90% confidence levels in the ranges we give) and the tools to help us in training our team to become expert in developing 90% confidence levels in the ranges for variables in our case models. Appendices in the book provide surveys for our use in getting across the idea of developing ranges. A “final test” I enjoy and now often use is to imagine a “wheel-of-fortune” type wheel with 9-10ths (90%) of the surface area covered in a color representing “You win $1,000.” The remaining 10% of the wheel is “You win $0.” Now, play the game – would you rather accept a bet on winning $1,000 if the increased sales actually come within the range you gave, or would you rather “spin the wheel”? The only correct answer for a properly calibrated range estimate is “it doesn’t matter” because I have a 90% chance of winning either way.

As a companion to this very well written book, there is a web site that provides additional information and examples. How to Measure Anything should definitely be on every business owner’s and manager’s desk.

Go to Part II of this review – Monte Carlo Sample.