Ten Fully Worked Guesstimates With Indian Numbers
Most candidates think guesstimates are about knowing the “right” number. They are not. A strong guesstimate is a clean story of assumptions, drivers, arithmetic and sanity checks - good enough for business judgment when perfect data is unavailable.
- A guesstimate is a structured approximation: define scope, break into drivers, assume round numbers, calculate, then sanity-check.
- Always freeze the unit, geography, time period and segment before touching numbers.
- Use either top-down logic, starting from population or market, or bottom-up logic, starting from capacity or transactions.
- Round deliberately. In interviews, 140 crore, 2 crore, 10 lakh and 30 days are acceptable working assumptions if stated clearly.
- The answer is judged less on precision and more on structure, realism, communication and correction when challenged.
- Every answer should end with: “So my estimate is X, likely within a range of Y to Z because of these two sensitive assumptions.”
Big Picture: The Five-Move Guesstimate Machine
A guesstimate is not a guessing contest. It is a small business model built in public. You show how demand, supply, frequency, price, capacity or penetration combine to create a believable answer.
The Core Method Before the Ten Examples
Before you practise the numbers, lock the method. Most weak answers fail in the first 30 seconds because the candidate starts calculating before defining the problem. If you need a deeper pre-skill, revise defining the problem before solving it because guesstimates punish vague problem statements.
Definitions You Can Say in One Breath
- Guesstimate: A structured, assumption-led approximation used when exact data is unavailable or unnecessary.
- Driver tree: A breakdown of an outcome into causal variables that multiply or add up to the estimate.
- Top-down estimate: Starting from a large base, such as population, then narrowing through filters.
- Bottom-up estimate: Starting from operational units, such as stores, seats, riders or machines, then scaling up.
- Sanity check: A quick independent test that checks whether the answer is directionally believable.
Top-Down or Bottom-Up: Which Route Should You Pick?
If the question is about demand, start top-down. If it is about capacity, start bottom-up. If you can do both quickly, use one as the main answer and the other as a sanity check.
Ten Fully Worked Indian Guesstimates
All numbers below are interview assumptions, not official market data. The point is not to memorise them. The point is to practise clean decomposition, Indian units and sanity checks.
1. Paid cups of chai sold in Mumbai in one day
- Scope: Paid chai cups sold outside the home in Mumbai per day.
- Assume population: 2 crore.
- Adults: 60% of population = 1.2 crore.
- Chai drinkers: 70% of adults = 84 lakh.
- Buy chai outside: 40% on an average day = 33.6 lakh people.
- Cups per outside buyer: 1.2 cups.
- Calculation: 33.6 lakh × 1.2 = 40.3 lakh.
- Answer: Around 40 lakh paid chai cups per day.
- Sanity check: If a tea stall sells 300 cups/day, 40 lakh cups implies roughly 13,000 equivalent stalls, including offices, canteens and roadside sellers. That feels directionally plausible for a dense city.
2. UPI payments at a large MBA college canteen in one day
- Scope: Number of UPI transactions at one campus canteen per day.
- Students on campus: 3,000.
- Visit canteen daily: 70% = 2,100 students.
- Average purchases per visitor: 1.2.
- Total purchases: 2,100 × 1.2 = 2,520.
- Share paid through UPI: 90%.
- Calculation: 2,520 × 90% = 2,268.
- Answer: Around 2,300 UPI payments per day.
- Sanity check: If peak hours generate 150 to 200 transactions per hour and off-peak hours generate fewer, a 2,000-plus daily total is believable.
3. Food-delivery biryani orders in Bengaluru in one day
- Scope: Biryani orders placed through food-delivery apps in Bengaluru per day.
- Population: Assume 1.3 crore.
- App-addressable population: 30% = 39 lakh.
- People ordering any food today: 20% of addressable = 7.8 lakh.
- Biryani share of food orders: 12%.
- Calculation: 7.8 lakh × 12% = 93,600.
- Answer: Around 90,000 to 1 lakh biryani orders per day.
- Sanity check: The estimate is sensitive to daily ordering penetration. If only 15% order today, the answer falls to about 70,000.
4. Monthly revenue of a 3-screen multiplex in Pune
- Scope: Ticket revenue only, excluding food and beverages.
- Seats per screen: 180.
- Total seats: 3 × 180 = 540.
- Shows per screen per day: 4, so total shows = 12.
- Daily seat capacity: 540 × 4 = 2,160 seats.
- Average occupancy: 35% across weekdays and weekends.
- Tickets sold per day: 2,160 × 35% = 756.
- Average ticket price: ₹250.
- Daily revenue: 756 × ₹250 = ₹1,89,000.
- Monthly revenue: ₹1,89,000 × 30 = ₹56.7 lakh.
- Answer: Around ₹55-60 lakh per month in ticket revenue.
- Sanity check: At 50% occupancy, revenue would rise to about ₹81 lakh, so movie slate quality is the biggest swing factor.
5. Delivery riders needed for a quick-commerce dark-store cluster
- Scope: Riders needed for one dense urban cluster during daily operations, not citywide.
- Orders per day: Assume 2,400.
- Orders per rider per hour: 2.5, allowing travel, handover and waiting.
- Active hours per rider: 8 hours.
- Orders per rider per day: 2.5 × 8 = 20.
- Base riders needed: 2,400 ÷ 20 = 120.
- Buffer for absence and peaks: 20%.
- Calculation: 120 × 1.2 = 144.
- Answer: Around 145 riders for the cluster.
- Sanity check: If each rider can do only 16 orders/day due to traffic, requirement rises to 180 riders. Capacity assumptions dominate this estimate.
6. Toothpaste tubes sold in India in one year
- Scope: Consumer toothpaste tubes sold annually, all pack sizes converted into tube count.
- Population: Assume 140 crore.
- Toothpaste users: 75% = 105 crore.
- Average tube sharing: One tube used by 2 people.
- User-equivalent households: 105 crore ÷ 2 = 52.5 crore tube-using pairs.
- Tubes per pair per year: 6.
- Calculation: 52.5 crore × 6 = 315 crore tubes.
- Answer: Around 300-320 crore toothpaste tubes per year.
- Sanity check: If average usage is 4 tubes per pair instead of 6, the answer drops to 210 crore, so consumption frequency is the key sensitivity.
7. Daily transactions at a busy Delhi Metro station
- Scope: Passenger entry transactions at one busy station in a day.
- Catchment working population: 2 lakh people.
- Use metro on a typical day: 25% = 50,000 people.
- Use this station as entry: 60% = 30,000 entries.
- Non-work or occasional users: Add 20% = 6,000.
- Calculation: 30,000 + 6,000 = 36,000.
- Answer: Around 35,000-40,000 entry transactions per day.
- Sanity check: Spread over 16 operating hours, that averages about 2,250 entries/hour, with much higher morning and evening peaks.
8. Annual market size for premium gym memberships in Gurugram
- Scope: Annual membership revenue from premium gyms, not budget gyms or personal training add-ons.
- Urban addressable adults: Assume 10 lakh.
- Can afford premium gym: 25% = 2.5 lakh.
- Interested in fitness membership: 20% = 50,000.
- Actually buy premium membership: 50% = 25,000 members.
- Average annual fee: ₹30,000.
- Calculation: 25,000 × ₹30,000 = ₹75 crore.
- Answer: Around ₹75 crore per year.
- Sanity check: If 50 premium gyms average 500 members each, that also gives 25,000 members. The bottom-up check supports the top-down answer.
9. ATMs needed in a tier-2 city
- Scope: Number of ATMs required to serve cash withdrawals in one city.
- Population: 20 lakh.
- Adults with bank access: 60% = 12 lakh.
- Use ATM at least once a month: 50% = 6 lakh users.
- Average ATM visits per user per month: 2.
- Monthly ATM transactions: 6 lakh × 2 = 12 lakh.
- Transactions one ATM can handle per month: 200/day × 30 = 6,000.
- Base ATMs needed: 12 lakh ÷ 6,000 = 200.
- Buffer for distribution and downtime: 20%.
- Answer: Around 240 ATMs.
- Sanity check: This is not only a demand problem. ATMs must be geographically distributed, so low-utilization areas still need coverage.
10. Annual revenue from airport cab rides in Hyderabad
- Scope: Cab rides from airport to city and city to airport in one year.
- Passengers per day: Assume 80,000 airport passengers.
- Need ground transport: Nearly all, but cab share assumed at 35% = 28,000 passengers.
- Average passengers per cab: 1.6.
- Cab rides per day: 28,000 ÷ 1.6 = 17,500 rides.
- Average fare: ₹900.
- Daily cab revenue: 17,500 × ₹900 = ₹1.575 crore.
- Annual revenue: ₹1.575 crore × 365 ≈ ₹575 crore.
- Answer: Around ₹550-600 crore per year.
- Sanity check: Cab share and average fare drive the estimate. If cab share is 25%, the answer falls closer to ₹410 crore.
Mini Case Study: Urban Company and the Supply-Side Guesstimate
Urban Company shows why guesstimates are not classroom puzzles: a marketplace must estimate demand, partner capacity and service reliability before scaling a local category.
Imagine Urban Company considering a new salon-at-home micro-market in a dense Indian city. The real business problem is not “how many customers exist?” It is: how many trained professionals should be onboarded so customers get fast slots without partners sitting idle?
Here is an interview-style reconstruction using assumptions, not company-disclosed numbers.
The estimate says the launch needs roughly 115 professionals. The primary driver is partner capacity per month, because small changes in jobs per day or utilization quickly change supply required. Supporting drivers include target household density, monthly booking penetration, repeat frequency and buffer for cancellations or peak-hour demand.
Lesson: The best guesstimates are decision tools. Here, the answer is not just “115 professionals.” It tells the business what to test first: utilization, booking frequency and peak-time availability.
How AI Changes Guesstimates in 2026
AI does not remove the need to estimate. It changes how you practise, benchmark and pressure-test your assumptions.
- Faster assumption benchmarking: Tools like Perplexity can help you quickly understand category context, but in an interview you must still state assumptions as your own working numbers.
- Cleaner driver trees: ChatGPT or Claude can generate alternate decompositions - population-led, capacity-led, transaction-led - so you learn multiple routes for the same question.
- Sharper sensitivity testing: AI can vary one assumption at a time and show which driver changes the answer most, which is exactly how strong candidates discuss uncertainty.
Use ChatGPT as a mock interviewer: paste one of the ten guesstimates above, ask it to challenge three assumptions, then answer aloud in two minutes. For a structured practice routine, use practising cases with AI as a mock interviewer.
Interview Relevance
“Estimate the number of food-delivery orders placed in Bengaluru in a day. Walk me through your assumptions and tell me where your estimate could be wrong.”
If your first structure is weak, pause and rebuild it. Interviewers forgive corrected assumptions faster than they forgive confident chaos.
Common Mistake
The biggest mistake is calculating before defining the unit. Candidates estimate “market size” without clarifying whether it means users, orders, revenue, monthly demand or annual demand. The fix: start every answer with one sentence - “I will estimate X, in Y geography, over Z time period, for this customer segment.”