Guesstimates interview questions

What a strong answer covers

• Clarify the question: define 'piano tuner' as a professional who tunes pianos, and 'Chicago' as the city proper or metro area.
• Estimate the population of Chicago (e.g., 3 million) and the number of households (e.g., 1.2 million).
• Assume a percentage of households with a piano (e.g., 10%) and add institutions like schools, churches, and concert halls.
• Estimate how often a piano needs tuning (e.g., once per year) and how many tunings a tuner can do per day (e.g., 3-4).
• Calculate: total pianos / (tunings per tuner per year) to get the number of tuners, and present the result with a sanity check.

Where people lose the point

• ×Forgetting to include institutional pianos (schools, churches) which can significantly increase the count.
• ×Using an unrealistic percentage of households with pianos without justification.
• ×Failing to state assumptions clearly or jumping straight to a number without showing the breakdown.

What a strong answer covers

• Clarify whether 'consumed' includes home-brewed and office coffee or only purchased from cafes; state your assumption.
• Estimate NYC population (8.5 million) and the percentage of adults who drink coffee daily (e.g., 60-70%).
• Segment by frequency: some drink 1 cup, others 2-3 cups; assume an average of 2 cups per coffee drinker per day.
• Add a factor for tourists and commuters who may consume additional cups while in the city.
• Calculate: population * coffee drinker fraction * average cups per drinker, and adjust for non-adult population.

Where people lose the point

• ×Using the entire population including children without adjusting for the fact that many don't drink coffee.
• ×Forgetting to account for multiple cups per person per day.
• ×Not clarifying the scope (home vs. purchased) and giving a single number without context.

What a strong answer covers

• Estimate the number of cars in the US (e.g., 250 million) and the average number of cars served per gas station per day.
• Assume an average driver fills up once a week, and a gas station serves a certain number of customers per day (e.g., 300).
• Calculate total fill-ups per week: cars * 1 fill-up/week = 250 million fill-ups/week, which is about 35.7 million per day.
• Divide daily fill-ups by average customers per station per day (300) to get roughly 119,000 stations.
• Cross-check with known data: there are about 150,000 gas stations in the US, so your estimate should be in that ballpark.

Where people lose the point

• ×Using the number of cars but forgetting that some cars are not driven weekly or are electric.
• ×Overestimating the number of customers a station can serve per day.
• ×Not cross-checking the final number with a known reference point.

What a strong answer covers

• Segment the population by gender and typical haircut frequency: men every 4-6 weeks, women every 6-8 weeks.
• Assume 50% male and 50% female, and that children get haircuts less frequently (e.g., every 8 weeks).
• Calculate average haircuts per person per year: men ~10, women ~7, children ~6, then weight by population share.
• Multiply the average per person by the total population (1 million) to get total haircuts per year.
• Present the result and note that it includes both salon and home haircuts, which may affect the estimate.

Where people lose the point

• ×Using a single average frequency for everyone without segmenting by gender or age.
• ×Forgetting to include children or assuming they get haircuts as often as adults.
• ×Not stating whether the estimate includes home haircuts or only professional ones.

What a strong answer covers

• Estimate the US population (330 million) and the percentage that owns a smartphone (e.g., 85%).
• Determine the average replacement cycle for a smartphone (e.g., every 3 years).
• Calculate annual sales: smartphone owners / replacement cycle = (330M * 0.85) / 3 ≈ 93.5 million.
• Adjust for first-time buyers (e.g., teenagers getting their first phone) and business purchases.
• Cross-check with known industry data: annual US smartphone sales are around 150-170 million, so your estimate should be in that range.

Where people lose the point

• ×Using the entire population without excluding children who don't own phones.
• ×Forgetting that some people upgrade more frequently than every 3 years, which increases sales.
• ×Not considering that some phones are sold to businesses or as replacements for lost/broken devices.

What a strong answer covers

• Estimate the number of buildings in NYC (e.g., 1 million) and the percentage that have elevators (e.g., 10%).
• Segment buildings by type: residential high-rises, office buildings, hotels, and commercial buildings.
• Assume an average number of elevators per building that has them (e.g., 2-3 for residential, 5-10 for commercial).
• Calculate: number of buildings with elevators * average elevators per building.
• Cross-check with known data: NYC has about 70,000 elevators, so your estimate should be in that ballpark.

Where people lose the point

• ×Assuming all buildings have elevators, when many walk-ups do not.
• ×Using the total number of buildings without segmenting by height or type.
• ×Forgetting that some buildings have multiple elevators, especially high-rises.

What a strong answer covers

• Clarify whether you mean black cabs or all ride-hailing vehicles (Uber, etc.); state your assumption.
• Estimate London's population (9 million) and the number of taxi trips per day (e.g., 300,000 for black cabs).
• Assume an average number of trips per taxi per day (e.g., 20-30).
• Divide total daily trips by trips per taxi per day to get the number of taxis.
• Cross-check with known data: there are about 21,000 black cabs in London, so your estimate should be close.

Where people lose the point

• ×Confusing black cabs with private hire vehicles (minicabs) which are much more numerous.
• ×Using the entire population as potential riders without considering that many use public transport.
• ×Not stating whether the estimate includes only licensed taxis or also ride-hailing services.

What a strong answer covers

• Estimate the US population (330 million) and the percentage that buys at least one book per year (e.g., 70%).
• Segment by frequency: some buy 1-2 books, others 10+; assume an average of 5 books per buyer per year.
• Calculate: population * buyer fraction * average books per buyer = 330M * 0.7 * 5 ≈ 1.155 billion.
• Adjust for e-books and audiobooks, which are also 'books' but may have different purchase patterns.
• Cross-check with industry data: US book sales are around 700-800 million print copies, so your estimate should be in that range.

Where people lose the point

• ×Using the entire population without excluding non-readers or children who may not buy books.
• ×Forgetting that many books are bought as gifts or for educational purposes, increasing the count.
• ×Not specifying whether the estimate includes e-books and audiobooks.

What a strong answer covers

• Estimate San Francisco's population (850,000) and the percentage of households without in-unit laundry (e.g., 40%).
• Assume an average household size (e.g., 2.5 people) to get the number of households (340,000).
• Calculate the number of households that need laundromats: 340,000 * 0.4 = 136,000 households.
• Assume each laundromat serves a certain number of households (e.g., 2,000) based on capacity and location.
• Divide households needing service by households per laundromat to get the number of laundromats (e.g., 68).

Where people lose the point

• ×Using the total population instead of households.
• ×Assuming all households have in-unit laundry, which is not true in dense cities.
• ×Not considering that some people use laundry services or wash at friends' homes.

What a strong answer covers

• Estimate the number of battery-powered devices per household that use AA batteries (e.g., remote controls, toys, flashlights).
• Assume an average of 4 devices per household, each using 2-4 batteries, and a replacement frequency of once a year.
• Calculate total batteries per household per year: 4 devices * 3 batteries * 1 replacement = 12 batteries.
• Multiply by the number of households in the US (e.g., 130 million) to get 1.56 billion batteries.
• Cross-check with industry data: US sales of AA batteries are around 3 billion per year, so your estimate should be in that range.

Where people lose the point

• ×Forgetting that many devices use rechargeable batteries, reducing the need for disposables.
• ×Using the population instead of households.
• ×Not accounting for commercial use (e.g., in businesses, hospitals) which adds to the total.

What a strong answer covers

• Estimate NYC's population (8.5 million) and the number of meals eaten out per person per week (e.g., 3).
• Calculate total restaurant meals per week: 8.5M * 3 = 25.5 million meals/week.
• Assume an average restaurant serves a certain number of meals per week (e.g., 1,000 based on seating and turnover).
• Divide total meals by meals per restaurant per week to get the number of restaurants.
• Cross-check with known data: NYC has about 24,000 restaurants, so your estimate should be in that ballpark.

Where people lose the point

• ×Using the entire population without considering that many meals are eaten at home or at work cafeterias.
• ×Overestimating the number of meals per restaurant per week.
• ×Not distinguishing between full-service restaurants, fast food, and food trucks.

What a strong answer covers

• Estimate Amsterdam's population (850,000) and the percentage of residents who own a bicycle (e.g., 70-80%).
• Calculate the number of bicycle owners: 850,000 * 0.75 ≈ 637,500.
• Add an estimate for households with multiple bikes (e.g., some people own more than one) and rental bikes.
• Assume an average of 1.2 bikes per owner to account for multiple bikes.
• Present the final estimate and note that Amsterdam is known for cycling, so the percentage is high.

Where people lose the point

• ×Using a low ownership percentage typical of car-centric cities.
• ×Forgetting to include rental bikes and bikes owned by businesses.
• ×Not considering that some residents may not own a bike but use shared bikes.

What a strong answer covers

• State that the NYC Marathon is one of the largest in the world, with a known cap of about 50,000 participants.
• Assume that a certain percentage of starters finish (e.g., 95-98%).
• Calculate finishers: 50,000 * 0.97 ≈ 48,500.
• Mention that the actual number is around 50,000 finishers, so your estimate should be close.
• Note that the number can vary slightly due to weather and other factors.

Where people lose the point

• ×Guessing a number without any basis, such as 10,000.
• ×Forgetting that not all starters finish the race.
• ×Not using known facts about the marathon's size.

What a strong answer covers

• Clarify the definition of 'airport' – include all paved runways or only commercial airports? State your assumption.
• Estimate the number of countries (195) and the average number of airports per country (e.g., 50 for smaller countries, 500 for the US).
• Use a top-down approach: total airports ≈ 195 * average per country, but adjust for outliers like the US with ~5,000 public airports.
• Alternatively, use known data: there are about 40,000 airports worldwide, with over 20,000 in the US alone.
• Present a reasoned estimate and cross-check with known figures.

Where people lose the point

• ×Using only commercial airports, which are a small fraction of total airports.
• ×Assuming every country has a similar number of airports.
• ×Not clarifying the definition and giving a vague answer.

What a strong answer covers

• Estimate the US population (330 million) and the percentage that owns a smartphone (e.g., 85%).
• Estimate the market share of iPhones among smartphone users (e.g., 50%).
• Calculate: 330M * 0.85 * 0.50 ≈ 140 million iPhone users.
• Adjust for the fact that some people own multiple phones or that children may have iPhones.
• Cross-check with known data: there are about 120-150 million iPhone users in the US.

Where people lose the point

• ×Using the entire population without excluding non-smartphone users.
• ×Assuming a market share that is too high or too low without justification.
• ×Not considering that some people have both a personal and work iPhone.

What a strong answer covers

• Estimate the US population (330 million) and the average number of pizza slices consumed per person per year (e.g., 46 slices, a known statistic).
• Convert slices to pizzas: assume a large pizza has 8 slices, so 46 slices ≈ 5.75 pizzas per person per year.
• Multiply by population: 330M * 5.75 ≈ 1.9 billion pizzas per year.
• Adjust for the fact that not everyone eats pizza, but some eat more than average.
• Cross-check with industry data: about 3 billion pizzas are sold in the US each year, so your estimate should be in that range.

Where people lose the point

• ×Using a per-person average without considering that children and adults have different consumption.
• ×Forgetting to convert slices to pizzas correctly.
• ×Not cross-checking with known industry statistics.

What a strong answer covers

• Estimate the number of major hotels on the Strip (e.g., 30) and their average size (e.g., 3,000 rooms).
• Calculate rooms on the Strip: 30 * 3,000 = 90,000 rooms.
• Add rooms in downtown and off-Strip hotels, which might add another 30,000-40,000 rooms.
• Total estimate: 120,000-130,000 hotel rooms.
• Cross-check with known data: Las Vegas has about 150,000 hotel rooms, so your estimate should be close.

Where people lose the point

• ×Forgetting that many hotels are off the Strip and in downtown.
• ×Using an average room count that is too low for mega-resorts.
• ×Not considering that some hotels have over 5,000 rooms.