Much of my operating work has been in markets the large platforms underserve: an electric cab service in Jammu, a matrimony platform, a construction equipment marketplace, and restaurants and a magazine through my agency Cairros. The businesses had little in common. The markets they sold into behaved alike, and this is my attempt to write that behaviour down as a thesis I would invest against.
Where the demand is coming from
A report covered by The Week in February 2025 expects shoppers from tier-2 and tier-3 markets to drive a fourfold growth in Indian e-commerce by 2035. Redseer found that about 65% of users on Indian short-video platforms come from tier-2 cities and beyond. More of the new demand for consumer businesses is coming from smaller cities, and a lot of it is already online.
Demand being online does not mean supply is. The services people in these cities buy most, a ride, a wedding vendor, a machine for a site, are still delivered by small local operators who are found through a phone call. That mismatch is where I think the opportunity sits.
The incumbent is a relationship
In Jammu the main competitor for ZenCabs was a driver's number saved in a rider's phone. In matchmaking it is a relative or a family contact, and in construction equipment an owner the site manager already knows. These incumbents cost nothing to use and arrive with trust already built.
A new business has to beat them on something a relationship cannot promise, and in my experience that is reliability. A relationship is only as dependable as the one person behind it being free today. A service that turns up every time can win against one that turns up most of the time.
Reliability beats price
The instinct in a price-sensitive market is to discount. At ZenCabs we chose to compete on reliability, because in a small market a ride that shows up is worth more than a cheaper one that might not. We cut driver-side cancellations from 40% to 6% and user cancellations from 200 to 80 a day, and 27% of users became monthly active riders.
Discounts also carry a cost that arrives later. My Push and Absorb framework calls it the expectation channel: a festival discount becomes the price people expect all year, and it is the hardest change to take back.
Supply is the moat
In each of these markets the constrained side was supply. At ZenCabs we built recruitment, training and incentives for driver partners from scratch before pushing for riders, and onboarded 43 of them. At Equip9 the owners listing machines needed a different message and funnel from the contractors renting them.
Supply built like this is slow and does not scale with ad spend. The slowness protects the business, because a competitor arriving later has to do the same local work again, city by city.
Trust travels by word of mouth
In smaller cities and close communities, people ask each other before they try something new, so the quality of delivery becomes the main acquisition channel. At Shaadi Mangalam, where a family buys a matrimony package once, referral is effectively the only repeat business there is. At Cairros, a reputation in one sector was what got me into the next, and the agency grew across four sectors that way.
What it means for unit economics
Prices are lower outside the metros and so are many costs, but density is lower as well. Fewer customers per square kilometre means a cab drives further empty between rides and a vendor serves fewer customers a week. The businesses that work here tend to combine a high value per transaction with high use of an expensive asset. ZenCabs earns ₹83K per car per month at 75% utilisation, and long trips to the airport or to Katra raise the value of a single booking. I broke those numbers down in the unit economics of an electric cab fleet.
Where I would invest, and where I would be careful
I would look for services with fragmented local supply and demand that is either frequent or high in value: mobility, home and repair services, equipment rental, weddings and events, and local logistics. The model that tends to work is an operator that organises supply and stands behind its quality. It is harder to build than a listings site, and for the same reason harder to copy.
I would be careful with models that depend on metro density, such as very fast delivery, where the economics rely on many orders from a small area. I would be equally careful with businesses that need large amounts of borrowed capital for assets. BluSmart, which ran an owned electric fleet and stopped operating in April 2025 after a SEBI order over the use of its loans, showed how much of that kind of business's risk sits in its financing.
The wedding industry breakdown and construction equipment breakdown go deeper on two of these sectors.
What would prove this thesis wrong
- Large platforms reaching reliable supply in smaller cities at metro cost, which would remove the local operator's edge.
- Zero-commission subscriptions spreading so far that organised supply can no longer recruit drivers, vendors or owners.
- Retention in smaller cities staying low even when the service is reliable, which would mean trust matters less than I think.
- Lower density making utilisation impossible to raise past a ceiling that breaks the economics.
Writing down what would change my mind before backing a thesis is the investing version of the stop rule in my framework. I would check this list against the numbers every six months.
