If you are interested in quickly answering basic questions, saving time and resources while reducing risk, read on. This post is for you.
In my last post, I wrote about how people often prefer to be precisely wrong than vaguely right. All too often, clean-looking datasets are prized over messy information that may actually speak to the truth that fuels smart decisions. Recall great decision-makers don’t have to go to math camp. However, it is important to have an intuition for numbers. The good news? Practicing the habit of approximation will help you build the intuition you need.
The Power of Approximation
My co-author Dr. Oded Netzer, Vice Dean of Research at Columbia Business School, once observed a group of MBA students present an idea for a new business: an eco-lodge in Guatemala. Taking advantage of increasing popularity of eco-tourism, this enterprise would invite travellers to be part of living on a Guatemalan working farm.
To estimate the annual revenue of this business, the students started at the top. They researched the size of the global eco-tourism market, narrowed that down to Central America, and whittled that down to Guatemala. Then, they estimated the number of tourists visiting Guatemala who would prefer farm-stays and multiplied that by one family’s booking cost. When it was all done, the student team came up with an impressive $32 million figure.
While the other professors grilled the students on the accuracy of each of the number they used in their top-down analysis, Oded focused on something else. Despite not knowing much about the tourism industry, his gut told him that $32 million was too good to be true. With some basic assumptions and arithmetic, he came up with a much different estimate.
Common sense dictated that the farm-stay’s annual revenue couldn’t exceed the cost of a night’s stay, multiplied by the lodging’s total number of rooms, again multiplied by most of the nights in a year. Simple, clear, logical. Given the accommodations and location, he assumed one night would never cost more than $300. He also knew that properties like this probably wouldn’t have more than 50 rooms. Assuming that each room was occupied for 250 days of the year, the revenue estimate ended up being $3.75 million annually. Not even close to the students’ $32 million number.
Oded resisted the temptation to get bogged down by handwringing over the exactness of the numbers in the students’ analysis. He focused on a simple estimate instead and exposed its fundamental flaws with simple, back-of-the-envelope math.
Learning How to Guesstimate: A Key Skill to Build
The writer A.A. Bell defines guesstimate as “better than a guess, but not as guaranteed as an estimate.” Guestimates shouldn’t be taken as gospel, but they are very useful. They help us make the most of little data, they provide insights into whether a figure is far too large or far too small, and most importantly, they help build our intuition muscles.
You may have heard of Enrico Fermi, an Italian physicist who was a master of making approximations with scant data. He developed the Fermi Method, which uses educated guesses that could inform how you approach a problem. He did this by employing guesstimation techniques, tackling complex problems by breaking them down into smaller, simpler components. I’ll demonstrate below.
Let’s move from Guatemalan farm-stays to a more universal topic: baby blankets. If you were a company providing this essential item for maternity wards across the United States, how would you go about estimating how many need to be made to satisfy their daily demand?
You won’t know the number of maternity wards across the country, or the number of babies born per day off the top of your head. The secret is to start with the relevant figures you do know, applying Fermi’s approach to break the calculation into smaller, more manageable components.
You know there are about 340 million people living in the U.S., and about half are female (170 million). We can also guess from a look around at our friends that the average family has about two children. Common wisdom suggests that American women live for an average of 80 years. That leaves us with the following:
Number of women in the U.S. = 170M
Babies delivered per woman = 2
Average lifetime of a woman in days = 80 years X 365 days
Number of deliveries per day = 170M x 2 / (80 x 365)
With some basic math, our guesstimate approximates to 11,640 babies delivered per day, and thus the need as many baby blankets. The actual number of deliveries per day in 2024 (the most recent year available), according to official statistics? About 9,940. Our guesstimate came in at roughly 15% larger, and that’s totally fine. The number we came up with is entirely sufficient for what we were looking for—a ballpark estimate to help us make a smart decision.
Why this Works
Of course, situations exist where exact numbers are essential. You wouldn’t want rough estimates when designing a vaccine or launching a space shuttle. But that’s the point. Many business decisions, especially in the initial stages, don’t necessitate numerical exactness.
Guesstimates and approximation thrive when you’re asking questions like: “Is this business model viable?” or “Does this revenue forecast look right?” Not only do they answer basic questions sufficiently, but they also save time and resources while reducing risk. Consider a sales forecast that, despite being based on numerous assumptions around units sold, prices, and conversion rates, presents an exact-seeming number rounded to the nearest thousandth decimal. That’s not only confusing but misleading too.
Like any skill, practice is key to success. I encourage business leaders to regularly ask for guesstimates. Doing so not only builds your teams approximation muscles, but crucially, helps you feel more comfortable with rougher figures. Both are key to agile, smarter decision-making.
Context Still Matters
Tight statistical confidence intervals may buttress your confidence in a decision, but there’s no reason to strive for them in every decision. In fact, there is no uniform required level of confidence independent of the decision at hand. Confidence levels come down to the decision’s context and the risk involved in being wrong.
Understanding these two factors (context and risk) comes from the intuition you’ve built and common sense. If you’re choosing between two versions of an ad, being wrong 1 out of 5 times likely won’t land you in hot water. On the other hand, when it comes to open heart surgery, being wrong 20 percent of the time is an intolerable safety risk. When faced with a decision that seems difficult, consider the stakes. Unless you work in an operating room, chances are that some basic facts, your intuition, and a decent level of confidence are all you need to move forward.
Inserting Intuition into Your Process
Too often, it’s assumed that smart decision-making hinges on statistical precision—but that’s simply not the case. Being slow yet exact enables disrupters who are able to leverage intuition to blow past you. This not about operating with a lack of certainty but getting comfortable with estimates is a crucial skill. So, practice your approximation skills whenever the context allows it. Explore the Fermi Method. You might be surprised at how an informed guess and some fifth-grade math open the door to a different dialogue.
Interested in learning more about how you can insert intuition into your decision-making process? Reach out. Let’s talk.
Christopher Frank, CEO, PSB Insights