How to solve guesstimates and market sizing questions?

Mohtashim
Analytics Vidhya
Published in
7 min readJan 14, 2020
Photo by Steve Halama on Unsplash

A couple of my colleagues were discussing these guesstimates questions. I didn’t even know if such a word did exist. But it turns out guesstimates are extremely popular among people lining up for consulting and analyst jobs and recruiters headhunting for consulting and analysts profiles.
The etymology of guesstimate is not strenuous to surmise. It is a blend of guesswork and estimation. Sounds fair!

Before prancing towards the method, one important thing that you should keep in mind is that never will the recruiter stress for an accurate answer. They are mostly trying to observe your approach. So you can take your assumptions freely as long as they are not too intuitive.

I will share a sample question, and we will work out a solution for it while discussing the method used.

Question: How much money a Burger King outlet makes
per day in Delhi?

We will solve the above question using the approach mentioned below.

The guesstimates questions can sometimes be extremely vague and ambiguous. It is always advisable to clarify all your doubts about the question. Take an example. You might be asked to find the number of flights that depart from Delhi. Now this question can have multiple interpretations. Is the interviewer concerned about passenger flights or cargo flights, or both? Is he talking about local flights or international flights? Is he only talking about IGI Airport or other airports in Delhi like Safdarjung, Hindon AF Station? All these questions need to be clarified before you start looking to answer the question. Once you get what the interviewer is looking for, it will be easy to assume the numbers and situations.

The second step, which is perhaps the most important step, is deciding which route to take. There are two possible ways to break down the problem. The first is the Top-Down approach. The top-down approach is useful when you have to start with the population of a city, state or country. Once you pick up a number that is close enough to the actual number (Hint: Make sure you know the population details of your city, state and country before you appear for the interview), the next step is to breakdown the population into age, gender, income groups etc. For example, if you get a question to find out the number of Gillette Shaving Foam users in Delhi, you will have to use a top-down approach. First, you need to start with the population of Delhi, and then you need to find out the number of males. The first bifurcation is on the gender parameter. However, it does not end there. Now you will break males into age groups. Males below 18 years old won’t be shaving. You can eliminate at least 25% of those males. The remaining males will be divided into two categories. Those who shave and those who trim. You can eliminate at least another 20% of the males who might be trimming. The next set of men will be those who would be shaving regularly. But not all of them would be using shaving foam. Some of them might be using shaving cream. So you need to break down further. And so on.

If you recall, we started the solution to the above problem by picking up the entire population of Delhi and eventually broke it down to the target audience of shaving foam. This method is essentially a top-down approach.

The next method is the Bottom-Up approach. This method is useful when you have to start the solution by assuming something that one individual in your target group would be doing. For example, you might want to assume how much one individual spends in a Mcdonald’s outlet in a day. Based on that average number, you can extrapolate it to the entire population and find out how much a McD outlet earns in a day. We will use the bottom-up approach in our example. So I am not explaining in detail right now.

As soon as you get the question, it would be best if you decided which method you will use to solve the question. Understand this point that guesstimates are not difficult because you have to make many assumptions; they are difficult because you have to eliminate all the grot information. Once you procure the method you will use, all the grot information that needs to be eliminated starts flooding in your head.

Now you know the method you will use and have broken the problem until you have reached your Target Group. All you have to do is solve all those pieces and consolidate them and present the answer. Let’s try to answer the question that I wrote above. Posting it below as well.

Question: How much money a Burger King outlet makes
per day in Delhi?

Take some time and find out which method you are going to use to solve this particular question.

There is no right or wrong answer in guesstimates. As mentioned above, what matters is your approach to the problem. My approach might differ from yours. Your approach might differ from the interviewer’s. As long as you can offer logical explanations for your reasons and eliminate all the grot information, you are on the right track.

Coming back to the above question. First, I will clarify the doubts that I have about this question.

Is he talking about any specific outlet because a CP outlet might earn more than some outlet present in East Delhi?
By ‘makes per day,’ is he talking about the revenue or profit?
Do I need to take any specific day because a store might earn more on weekends compared to weekdays?

After clarifications, I have to espouse the method that I will use to solve the question. The method that I am picking up is Bottom-Up.

Initial Assumptions:

  1. An average individual spends INR 200 in a Burger King Outlet.
  2. The sale is not affected by the weekend factor. It is uniform for the entire week.
  3. All the stores in Delhi have similar customer footfall.
  4. The Burger King outlet has 3 cash counters.
  5. The store opens at 10 AM and closes at 10 PM.

So I have taken these initial assumptions and have conveyed the same to the interviewer.

Solution:

An average individual spends Rs. 200 in a typical Burger King outlet. I am going to divide the day into 2 buckets. The first bucket is a low flow bucket meaning the customer footfall is low. And the second bucket is a high flow bucket meaning the customer footfall is extremely high.

Buckets:

10 AM — 12 PM: Low-Flow Bucket (2-hour window)

12 PM — 4 PM: High-Flow Bucket (4-hour window)

4 PM — 6 PM: Low-Flow Bucket (2-hour window)

6 PM — 10 PM: High-Flow Bucket (4-hour window)

The low flow bucket has been chosen for odd hours. The high-flow buckets are the lunch and dinner times.

Further assumptions.

  1. The average time is 3 minutes to place one order and is the same for high flow and low flow periods.
  2. The high flow has all the three counters utilized, and the time loss for the next set of customers is zero assuming there is a long queue.
  3. The time taken for the next set of customers to come into the store during a low flow period is 7 minutes.

Calculations:

For a high-flow period.

Since all three counters are utilized simultaneously, and it takes 3 minutes to place one order. So in 3 minutes, three orders are being placed (summation of the orders from the three counters).

And in 60 minutes, 60 orders are being placed (no loss of time as there is a long queue during the high flow period).

The average price for one order is Rs. 200. So in 60 minutes, 60*200=12,000 worth of money is coming into the cash bank.

One high flow period is for 4 hours; hence, the total money earned during that period is 4*12000 = INR 48,000.

Assuming similar performance in the second high flow period, another 48,000 is being earned. The total money earned during the high flow period is 2*48,000 = INR 96,000.

Since we assumed that the time loss between the outgoing and incoming customer is zero in the above case, this will be the maximum revenue that we will be earning during the high flow period.

For a low-flow period.

Still assuming the three counters are being utilized simultaneously. However, the customer flow is low, so the next set of customers arrive after 7 minutes. It is still a fair assumption to take because we have taken pretty high time difference between the two sets of customers. In the low-flow period, we are getting three orders in 3+7=10 minutes.

In 60 minutes, we are getting 3*6=18 orders during the low flow period. The average price of one order is Rs. 200. So in 60 minutes, 18*200=3,600 worth of money is coming into the cash bank.

One low flow period is for 2 hours; hence, the total money earned during that period is 2*3600 = INR 7,200.

Assuming similar performance in the second low flow period, another 7,200 is being earned. The total money earned during the low flow period is 2*7,200 = INR 14,400.

The total money earned during an entire day is 96000 + 14400 = INR 110,400.

This is how a bottom-up method is used to solve a guesstimate question. I have tried to make it as simple as possible. If you have a different solution, do mention it in the comments, I would love to read it.

Should you have any questions on a guesstimate, do mention it as well; I will try to solve it and upload the solution.

Good luck with your next interview!

--

--

Mohtashim
Analytics Vidhya

I don’t write as much as I read. Passionate about Data Science and Machine Learning. Loves teaching.