Market sizing: 24 questions with worked answers

1. Start from the population: about 68 million people.
2. Assumption: at home, one baguette feeds four people for a day, averaged over the whole population (some never buy one, others buy one a day).
3. Assumption: food service (sandwich shops, canteens, restaurants) adds 20% on top of household volume.

Calculation: 68 million ÷ 4 = 17 million, then 17 million × 1.2 = 20.4 million

Result: About 20 million baguettes a day

Sanity check: That is a little under one baguette per household per day across 30 million households, which is plausible.

Classic trap: Reasoning per household with one baguette a day for everyone: you forget households that buy no fresh bread, and you forget food service.

1. The purchase is decided at household level: start from about 30 million households.
2. Assumption: 2% of households buy an electric bike in a given year.
3. Assumption: the average price paid is 2,000 euros.

Calculation: 30 million × 2% = 600,000 bikes, then 600,000 × €2,000 = €1.2 billion

Result: About €1.2 billion a year

Sanity check: At that pace one household in five would own one after ten years: ambitious but not absurd for a product still being adopted.

Classic trap: Mixing up the installed base (bikes in use) and the flow (bikes sold this year). The question is about the flow.

1. Start from about 68 million people.
2. Assumption: 60% of the population wears prescription glasses, a low share among children and a very high one after 50.
3. Assumption: a wearer replaces their pair every three years.

Calculation: 68 million × 60% = 40.8 million wearers, then 40.8 million ÷ 3 = 13.6 million

Result: About 13.6 million pairs a year

Sanity check: That is one pair for every five people each year, consistent with a product worn by a majority and kept for several years.

Classic trap: Forgetting the replacement cycle and answering with the number of wearers, which is a base and not annual sales.

1. A cat belongs to a household: start from about 30 million households.
2. Assumption: 30% of households have at least one cat, and those households have 1.5 on average.
3. Assumption: a cat costs 15 euros of food a month, so 180 euros a year.

Calculation: 30 million × 30% × 1.5 = 13.5 million cats, then 13.5 million × €180 = €2.43 billion

Result: About €2.4 billion a year

Sanity check: Spread over 30 million households, that is about 80 euros per household per year, a credible figure for a cost only a third of homes carry.

Classic trap: Counting one cat per cat-owning household. Many homes have two, and leaving that out removes a third of the market.

1. Start from about 68 million people.
2. Assumption: 75% are between 15 and 70, the age range that can sign up.
3. Assumption: 12% of that group holds an active membership.
4. Assumption: the average membership costs 30 euros a month, so 360 euros a year.

Calculation: 68 million × 75% = 51 million, 51 million × 12% = 6.12 million members, 6.12 million × €360 = €2.2 billion

Result: About €2.2 billion a year

Sanity check: Six million members is a little under one person in ten: enough that everyone knows a few, not so many that it is the norm.

Classic trap: Applying the membership rate to the whole population, including children and the very old.

1. A capsule machine belongs to a household: start from about 30 million households.
2. Assumption: 40% of households own a capsule machine.
3. Assumption: an equipped household uses 2 capsules a day on average over the year, holidays included.

Calculation: 30 million × 40% = 12 million households, 12 million × 2 = 24 million a day, 24 million × 365 = 8.76 billion

Result: About 8.8 billion capsules a year

Sanity check: At an assumed 35 cents a capsule, that would be a market of roughly 3 billion euros, about a hundred euros per French household.

Classic trap: Starting from the number of coffee drinkers without isolating how it is brewed: filter, beans and instant use no capsules.

1. Start from about 30 million households.
2. Assumption: a household owns 2 mattresses on average, counting children's beds and spare beds.
3. Assumption: a mattress is replaced every 12 years.
4. Assumption: hotels, residences and hospitals add 10% on top of household volume.

Calculation: 30 million × 2 = 60 million mattresses, 60 million ÷ 12 = 5 million, 5 million × 1.1 = 5.5 million

Result: About 5.5 million mattresses a year

Sanity check: That is roughly one household in six buying a mattress in a given year, which fits a rare purchase that everyone ends up making.

Classic trap: Counting one mattress per person: couples share theirs, and you overstate the base by a good third.

1. A subscription is shared within the home: start from about 30 million households.
2. Assumption: 60% of households pay for at least one service.
3. Assumption: those households hold 1.5 subscriptions on average, at 10 euros a month each.

Calculation: 30 million × 60% × 1.5 = 27 million subscriptions, then 27 million × €120 = €3.24 billion

Result: About €3.2 billion a year

Sanity check: A subscribing household would spend 180 euros a year, or 15 euros a month: the price of a service and a half, so the numbers hold together.

Classic trap: Counting users instead of paying subscriptions. An account shared by four people only pays once.

1. Assumption: out of 68 million people, about 28 million are in work.
2. Assumption: half of them work on premises where the employer provides coffee.
3. Assumption: 2 coffees a day over 210 working days, at a cost of 25 cents per coffee for the company.

Calculation: 28 million × 50% = 14 million employees, 2 × 210 × €0.25 = €105 per employee, 14 million × €105 = €1.47 billion

Result: About €1.5 billion a year

Sanity check: A 100-person company would spend a little over 10,000 euros a year on coffee, a figure an office manager would recognise.

Classic trap: Valuing the coffee at coffee-shop prices. The company buys beans or capsules in bulk, so the unit cost is far lower.

1. Assumption: about 28 million people work in France.
2. Assumption: 45% of them are given a laptop by their employer.
3. Assumption: a business laptop is replaced every 4 years.

Calculation: 28 million × 45% = 12.6 million laptops, then 12.6 million ÷ 4 = 3.15 million

Result: About 3.2 million laptops a year

Sanity check: At an assumed 1,000 euros each, the market would be a little over 3 billion euros, about a hundred euros per worker per year.

Classic trap: Forgetting that many jobs come with no dedicated computer: retail, restaurants, construction, care.

1. Assumption: about 28 million people work in France.
2. Assumption: 20% receive meal vouchers, the rest having a canteen, an allowance or nothing.
3. Assumption: a recipient gets 200 vouchers a year, worth 9 euros each.

Calculation: 28 million × 20% = 5.6 million recipients, 200 × €9 = €1,800, 5.6 million × €1,800 = €10.08 billion

Result: About €10 billion a year

Sanity check: The issuer earns a commission on that volume. At an assumed 3%, that is 300 million euros: the issuer's market is far smaller than the volume issued.

Classic trap: Answering without saying what you measure. Volume issued and issuer revenue differ by a factor of thirty.

1. First estimate the number of stores. Assumption: one supermarket for every 6,000 people.
2. With 68 million people, that gives about 11,300 stores.
3. Assumption: a store receives 15 pallets a day on average, between the small city-centre store and the hypermarket.

Calculation: 68 million ÷ 6,000 = 11,333 stores, then 11,333 × 15 = 170,000 pallets

Result: About 170,000 pallets a day

Sanity check: A semi-trailer carries about thirty pallets: that would take nearly 6,000 trucks a day, one truck for every two stores, which sounds reasonable.

Classic trap: Trying to go through household consumption in kilos. The supply-side route, store by store, is shorter and safer here.

1. Assumption: of about 28 million people in work, 40% work in an office.
2. Assumption: each desk takes up 15 square metres, shared areas included.
3. Assumption: cleaning costs 20 euros per square metre per year.

Calculation: 28 million × 40% = 11.2 million desks, 11.2 million × 15 sq m = 168 million sq m, 168 million × €20 = €3.36 billion

Result: About €3.4 billion a year

Sanity check: That is 300 euros per desk per year, a little over one euro per working day: plausible for a quick daily round.

Classic trap: Drifting out of scope. Factories, hospitals and shops get cleaned too, but the question is about offices.

1. Assumption: there is one point of sale for every 50 people (shops, restaurants, tradespeople, private practices).
2. With 68 million people, that gives 1.36 million points of sale.
3. Assumption: 85% accept cards, with 1.5 terminals per point of sale on average.

Calculation: 68 million ÷ 50 = 1.36 million, 1.36 million × 85% = 1.156 million, 1.156 million × 1.5 = 1.73 million

Result: About 1.7 million terminals

Sanity check: That is one terminal for every forty people. A shopping street with twenty shops would have about twenty-five, which matches what you see.

Classic trap: Counting one terminal per business. A supermarket has one per till, a restaurant passes several around the floor.

1. Assumption: France has about 20 million private-sector employees.
2. Assumption: half work in a company of fewer than 250 people, so 10 million payslips a month.
3. Assumption: the software is billed at 8 euros per payslip.

Calculation: 20 million × 50% = 10 million payslips a month, 10 million × €8 × 12 months = €960 million

Result: About €1 billion a year

Sanity check: A 50-person company would pay 4,800 euros a year, a small fraction of the cost of a payroll clerk: the price is bearable, so credible.

Classic trap: Mixing up the addressable market and the actual one. Some companies hand payroll to an accountant and buy no software at all.

1. Estimate the residential base. Assumption: 45% of the 30 million households live in a block of flats, and half of those blocks have a lift.
2. Assumption: one lift serves 20 flats, which gives 337,500 residential lifts.
3. Assumption: offices, hospitals, shops and stations add 40% to the base.
4. Assumption: a maintenance contract costs 2,500 euros per lift per year.

Calculation: 30 million × 45% × 50% ÷ 20 = 337,500, 337,500 × 1.4 = 472,500 lifts, 472,500 × €2,500 = €1.18 billion

Result: About €1.2 billion a year

Sanity check: For a block of 20 flats, maintenance would come to 125 euros per flat per year, an amount that would not look out of place in service charges.

Classic trap: Stopping at residential. Commercial buildings weigh heavily, and their lifts run far more, so they cost more to maintain.

1. Reason from the supply side, from what the shop can serve.
2. Assumption: 300 customers a day, concentrated in the morning, at lunch and in the late afternoon.
3. Assumption: an average basket of 4 euros, between a single baguette and a lunch deal.
4. Assumption: 310 opening days a year, with one closing day a week.

Calculation: 300 × €4 = €1,200 a day, then €1,200 × 310 = €372,000

Result: About €370,000 a year

Sanity check: Three hundred customers over twelve opening hours is one every two to three minutes on average, with queues at peak times: what you see outside a bakery that does well.

Classic trap: Using the price of a baguette as the average basket. Sandwiches, pastries and cakes make up most of the revenue.

1. Estimate demand, then divide it by what one hairdresser can handle.
2. Assumption: a resident goes to the hairdresser 5 times a year on average.
3. Assumption: a hairdresser handles 10 appointments a day, 230 days a year.

Calculation: 500,000 × 5 = 2.5 million appointments, 10 × 230 = 2,300 per hairdresser, 2.5 million ÷ 2,300 = 1,087

Result: About 1,100 hairdressers

Sanity check: That is one hairdresser for every 460 residents. With three per salon, the city would have about 360 salons, one for every 1,400 people.

Classic trap: Giving a number of salons when asked for hairdressers, or the reverse. Reread the question before you state your figure.

1. An order is placed per household. With 68 million people for 30 million households, that is about 2.3 people per household, so nearly 109,000 households in the city.
2. Assumption: 8% of households order a meal delivery on a Saturday night.
3. Assumption: 35% of those orders are pizza, with 1.8 pizzas per order.

Calculation: 250,000 ÷ 2.3 = 108,700 households, × 8% = 8,700 orders, × 35% = 3,040 pizza orders, × 1.8 = 5,480 pizzas

Result: About 5,500 pizzas

Sanity check: If a pizzeria turns out 150 delivery pizzas over the evening, it takes a little under forty outlets to serve the city, a credible number at that scale.

Classic trap: Reasoning per person and counting one pizza for each person ordering. The order belongs to a household, and its average size is under two pizzas.

1. Start from capacity. Assumption: one bay can wash 4 cars an hour, and the site is open 12 hours a day.
2. Assumption: the bays are busy 25% of the time on average, with weekend peaks and weekday lulls.
3. Assumption: a wash brings in 6 euros, and the site runs 360 days a year.

Calculation: 4 bays × 12 h × 4 washes × 25% = 48 washes a day, 48 × €6 = €288, €288 × 360 = €103,680

Result: About €100,000 a year

Sanity check: Forty-eight washes a day is four cars an hour across the whole site: you rarely see an empty car wash and rarely a queue, so the occupancy rate holds.

Classic trap: Working out maximum capacity and presenting it as revenue. Without an occupancy rate, you overstate by a factor of four.

1. Reason from train supply. Assumption: one departure every 30 minutes in each direction, from 6 am to 9 pm, so 30 departures each way.
2. Assumption: a train has 500 seats.
3. Assumption: the average load factor is 75%.

Calculation: 30 departures × 2 directions = 60 trains, 60 × 500 seats = 30,000 seats, 30,000 × 75% = 22,500 passengers

Result: About 22,500 passengers a day, both directions combined

Sanity check: Over a year, that is about 8 million trips. For a line linking two very large urban areas, the order of magnitude is defensible.

Classic trap: Going through demand (who has a reason to travel from Paris to Lyon?) when supply can be counted in three lines. Pick the approach whose assumptions you can defend.

1. Assumption: each screen has 200 seats and runs 5 showings a day.
2. Assumption: the average fill rate is 15%, between sold-out Saturday nights and nearly empty Tuesday mornings.
3. Assumption: a customer spends 10 euros, 7 on the ticket and 3 on snacks.

Calculation: 8 × 200 × 5 × 15% = 1,200 admissions a day, 1,200 × 365 = 438,000 admissions, 438,000 × €10 = €4.38 million

Result: About €4.4 million a year

Sanity check: If a local resident goes to the cinema three times a year, those 438,000 admissions imply a catchment of about 150,000 people, which matches a mid-sized city.

Classic trap: Using a 50% fill rate because the showings you attend are full. Most weekday showings are not.

1. Estimate demand in visits to the counter. Assumption: a resident visits a pharmacy 12 times a year on average.
2. Assumption: a pharmacy serves 150 customers a day, 300 days a year.

Calculation: 100,000 × 12 = 1.2 million visits, 150 × 300 = 45,000 per pharmacy, 1.2 million ÷ 45,000 = 26.7, rounded to 27

Result: About 27 pharmacies

Sanity check: That is one pharmacy for every 3,700 residents. In a city you pass one every few blocks, so the order of magnitude is right.

Classic trap: Forgetting that visits depend heavily on age. If the interviewer says the city is very old or very student-heavy, adjust the frequency instead of keeping your average.

1. With about 2.3 people per household, the city has nearly 130,000 households. Assumption: 0.9 cars per household in a city.
2. Assumption: 10% of cars are electric, and 30% of their owners cannot charge at home.
3. Assumption: those drivers charge 1.5 times a week at a public point, and one point handles 3 charges a day.

Calculation: 300,000 ÷ 2.3 × 0.9 = 117,400 cars, × 10% × 30% = 3,520 cars, × 1.5 ÷ 7 = 755 charges a day, ÷ 3 = 252 points

Result: About 250 charging points

Sanity check: That is one point for every fourteen electric cars without home charging. If the electric share triples, the network must triple: the result is very sensitive to that assumption, so say it.

Classic trap: Sizing for every electric car. Most charge at home or at work and almost never use on-street points.