At ZenCabs the fleet grew by 10+ cars a month to 40, and every batch of cars had to be justified by a market sizing and an expansion case. That case used our own data. This is a rebuild of the method from public numbers only, so anyone can check it, and so the assumptions are visible enough to argue with.
The question: how many app-booked cab trips could Jammu support in a month? I build it from three pools of demand, give each a low, base and high case, and then compare the result with what ZenCabs actually did.
The public inputs
| Input | Figure | Source |
|---|---|---|
| Jammu city population | 5,02,197 (2011) | Census of India |
| Jammu urban agglomeration | 6,57,314 (2011) | Census of India |
| Jammu airport passengers | 1.61 million (2024-25) | Ministry of Civil Aviation, in the Lok Sabha |
| Vaishno Devi pilgrims | 94.8 lakh (2024) | Shri Mata Vaishno Devi Shrine Board |
The Census is fifteen years old, so the first assumption is growth. At about 1.4% a year the 2011 urban agglomeration becomes roughly 8 lakh people in 2026. I use 8 lakh throughout. It matters less than the behavioural assumptions that follow.
Pool one: residents
Most residents will never book a cab through an app in a given month. They own a two-wheeler or a car, take an auto, or do not travel far. So the model starts from an addressable share: people with a smartphone, a reason to travel across the city and the income to pay a cab fare. My base case is 10% of 8 lakh, or 80,000 people, taking two app-booked trips a month each.
Base case: 80,000 people × 2 trips = 1,60,000 trips a month.
Pool two: the airport
The airport handled 1.61 million passengers in 2024-25, which counts arrivals and departures separately, so each passenger movement is one ground trip to or from the airport. That is about 1,34,000 a month. Many are collected by family or drive themselves; my base case is that 30% take a cab.
Base case: 1,34,000 × 30% = about 40,000 trips a month.
Pool three: pilgrims
94.8 lakh pilgrims visited the Vaishno Devi shrine in 2024, about 7.9 lakh a month, and Katra, the base town, is a road journey from Jammu. Many pilgrims take the train straight to Katra, travel by bus, or come in their own vehicles, so my base case is that 10% take a cab between Jammu and Katra. Pilgrims usually travel in groups, so I divide by an average of three people per cab.
Base case: 7.9 lakh × 10% ÷ 3 = about 26,000 trips a month.
The total, and how wrong it could be
| Pool | Low | Base | High |
|---|---|---|---|
| Residents | 5% × 1 trip = 40,000 | 10% × 2 trips = 1,60,000 | 15% × 3 trips = 3,60,000 |
| Airport | 20% = 26,800 | 30% = 40,250 | 40% = 53,700 |
| Pilgrims | 5% ÷ 3 = 13,200 | 10% ÷ 3 = 26,300 | 15% ÷ 3 = 39,500 |
| Trips a month | about 80,000 | about 2,27,000 | about 4,53,000 |
The range is wide, more than five times from low to high, and almost all of the width comes from one line: how many residents book cabs and how often. The airport and pilgrim pools together are between 40,000 and 93,000 trips a month in every case. That tells me where to spend research effort. A week of surveying residents would narrow the estimate more than a year of refining the airport numbers.
Checking it against reality
Six months in, ZenCabs ran 10,000+ rides a month at a ₹331 average order value. That is about 4.4% of the base case, 12.5% of the low case and about 2.2% of the high case. Even on the pessimistic view, the service had reached a small share of the market, which is what justified adding cars.
Priced at ZenCabs' ₹331 average order value, the base case is about ₹7.5 Cr of fares a month, or roughly ₹90 Cr a year. That understates it, because a Jammu to Katra trip costs far more than a city ride.
From market size to a fleet plan
A market size only becomes a decision when it turns into cars. At ZenCabs' actual rate of 250 rides per car a month (10,000 rides across 40 cars), serving 10% of the base case, about 22,700 trips, would need roughly 90 cars. That number is the useful output: it is a target the team can plan driver recruitment against.
The pace of adding cars should be set by utilisation. Each new batch is justified when the existing fleet is running near its target and cancellations from no car being available start to rise. If utilisation falls after a batch arrives, the next batch waits. That rule is a stop rule in the sense of my Push and Absorb framework: the condition for stopping is decided before the money is spent.
What would change the answer
- A resident survey. The resident pool drives most of the range, so this is the first thing to measure.
- Katra rail traffic. Direct trains reduce the pilgrims who pass through Jammu by road.
- A national aggregator arriving in force, which would split the market and raise driver costs.
- Seasonality. Pilgrim numbers rise sharply around festivals, so a monthly average hides peaks the fleet has to cover.
The economics of each car in that fleet are in the unit economics of an electric cab fleet.
