Paying Drivers to Wait Where Riders Have Not Arrived Yet
A ride hailing marketplace has to supply both sides at once in every city separately. Subsidy is the tool, and the economics only work where density arrives.
The Problem at Launch
A ride hailing platform is a marketplace with a specific difficulty. Its value to a rider depends on how quickly a car arrives, which depends on how many drivers are available nearby. Its value to a driver depends on how much time is spent carrying passengers rather than waiting, which depends on rider demand.
Neither side has a reason to join first. This is the cold start problem, and marketplaces solve it by paying one or both sides to participate before the market can sustain itself.
Subsidy in a marketplace is not a discount. It is the price of manufacturing liquidity that does not yet exist, and it has to be paid again in every new city.
Why the Market Is Local
The critical structural feature is that liquidity does not transfer between cities. A dense driver network in one metropolitan area does nothing for a rider in another. Each city is effectively a separate marketplace that must be built independently.
This distinguishes the model sharply from software businesses where a product built once serves every new customer at negligible additional cost. Here, expansion means repeating the subsidy in each new geography, which is why rapid growth and large losses coexisted in this industry for years.
Density Drives Everything
The variable that determines whether a city works is density: the concentration of trips within an area. It affects both sides simultaneously.
| Density effect | Consequence |
|---|---|
| Shorter distance to next rider | Less unpaid driver time |
| Higher driver utilisation | More earnings per hour worked |
| Faster pickup times | Better rider experience |
| More reliable service | Higher rider frequency |
The measure that captures this is utilisation, the share of a driver working time spent carrying a passenger. Unpaid time between trips is pure inefficiency in the system, and reducing it allows the platform to pay drivers more per hour while charging riders less per trip, because there is less waste between them.
This is why a dense urban market can reach sustainable economics while a sprawling low density one may never do so at any price.
The Take Rate
The platform earns a share of each fare, usually described as the take rate. Setting it is a continuous tension. Raise it and margin per trip improves while driver earnings fall, encouraging drivers to work elsewhere. Lower it and supply improves while unit economics weaken.
Because drivers can typically operate on multiple platforms simultaneously, supply is genuinely contestable in any city with more than one operator. That constrains how far take rates can be pushed and explains why competing platforms in the same city tend to compete away much of the available margin.
When Subsidies Stop
The strategic question underlying the entire model is whether subsidised demand persists once the subsidy ends. If riders used the service because it was artificially cheap, volume falls when prices normalise. If they used it because it was genuinely better than the alternatives, volume holds.
The honest answer varies by market and by use case. Where the service replaced an expensive or unreliable incumbent, retention has been strong. Where it competed against adequate public transport or personal vehicles on price alone, less so.
The industry shift toward prioritising contribution margin over growth reflected this question being asked seriously by investors, and the resulting changes were substantial: exiting unprofitable cities, raising prices, reducing driver incentives and focusing on the markets where density had actually been achieved.
Adjacent Demand
Adding food delivery and other services to the same driver network is an attempt to improve utilisation using assets already present. It fills demand troughs, since meal delivery peaks at different hours than commuting, and it spreads fixed platform costs across more transactions.
The complication is that the two businesses have different economics. Delivery involves lower fares, comparable driving time and often thinner margins, so combining them improves network utilisation without necessarily improving profitability per hour.
The Bottom Line
Ride hailing economics are decided city by city, because liquidity is local and must be purchased separately in each one. Density determines driver utilisation, which determines whether the platform can pay drivers acceptably and charge riders competitively while retaining a margin. The durable question for any such marketplace is what share of subsidised demand survives the removal of the subsidy, and the answer has proven to depend heavily on what the service was competing against.