New challenge: Fermi estimation

Introduction

During a case interview for a strategy consulting firm, I was asked to estimate a certain figure. Drawing on my usual competitive programming mindset, I mistakenly derived an upper bound and was satisfied with that result.

Realizing this was obviously not the right approach, I resolved to tackle new specialized fields with the appropriate mindset.

Fermi Estimation Practice - 01

I worked on the problem: “Estimate the annual market size of the domestic zipper industry.”

Although I had prepared in advance regarding methodology, when I actualy put it into practice, I found myself focusing too much on listing every possible factor without omission; I broke down the problem too strictly according to the domain, which led to a situation where I could no longer keep track of all the subdivided variables within the time limit - the critical problem I recognized. I expect that by anticipating the key variables in advance, I’ll be able to minimize the decomposition while still deriving a resonably convincing answer.

At the heart of Fermi estimation, it’s crucial to have a general sense of numerical values. Below is the metrics I gave up on decomposing simply because I wasn’t aware of them.

Population Distribution by Age Group in Japan as of the End of 2023

Age Group Figures as of End of 2023 Percentage
0–19yrs 115,961 13.0%
20–39yrs 195,197 21.9%
40–49yrs 215,798 24.2%
50–59yrs 209,900 23.5%
60yrs or older 154,713 17.4%
Aggregation 891,569 100.0%

Expected solution

  1. Define the market size as the annual sales revenue generated by zipper manufactures.
  2. Sales revenue can be calculated as the price per zipper multiplied by the number of zippers sold.
  3. Clothes are likely to be the dominant application of zippers, so I will estimate the market primarily based on clothing and use other applications as a supplementary adjustment.
  4. Japan has a population of approximately 125 million, which can be used as the basis for estimating the number of clothing items purchased annually.
  5. Assume that each person purchases approximately 10 pieces of clothing per year on average. This gives up 1.25 billion pieces of clothing per year.
  6. Assume that approximately 20% of clothing items use zippers. Therefore the number of clothing items using zippers every year is approximately 250 millions pieces.
  7. Assume that each zipper-equipped clothing item contains an average of 1.2 zipper units and an average price of zipper to be 100 yens, the clothing-related zipper market is about 30 billions.
  8. Zippers are also used in bags, shoes, luggage, bedding, outdoor equipment, and other products. Assume these applications account for approximately 50% of the clothing-related demand, the estimated annual domestic zipper market to domestic consumption is 45 billions.
  9. Finally, Japanese zipper manufactures also sell their products to overseas customers. Assume that exports account for approximately 900% of domestic-demand-based sales.
  10. Therefore, the estimated annual sales of the Japanese zipper industry would be approximately 4.5 trillion yens.

Uncertainties

  • other applications account for 50% of the zipper usage
  • 20% of clothing items using zippers
  • exports account for 900% of domestic-demand-based-sales