Insight, E-Mobility
A Lagos company is selling rides at roughly half the price of ride-hailing by grouping passengers who travel the same route at the same time. It did not build a better vehicle or a better algorithm. It removed the most expensive feature in urban mobility, which is spontaneity.
Nigeria's Shuttlers has launched Shuttlers Pod, a scheduled door-to-door shared car service that groups three or four passengers travelling similar routes (as reported). The company says scheduled trips can cost around half the price of conventional ride-hailing, and it is building on a network that already runs more than 430 buses and has completed over ten million journeys.
The timing is pointed, arriving after Uber's exit from the Nigerian market. But the interesting thing is not that a local company is replacing a departed one. It is that it is not replacing it with the same thing.
Strip a ride down to its costs and almost all of them are fixed to the vehicle and the hour rather than to the passenger. The car is depreciating whether it holds one person or four. The fuel difference between carrying one and four is small. The driver is paid for the hour either way. Insurance, maintenance and financing are all per vehicle.
Which means the single most powerful lever on the cost of a ride is the number of people in the car, and the second most powerful is the proportion of the driver's working day that is spent carrying anybody at all.
Conventional ride-hailing is structurally bad at both. It matches one request to one vehicle, so occupancy is usually one. And it requires the driver to be available and idle in order to be summonable, so a substantial share of the working day is spent circulating or waiting.
That model works where fares are high enough to carry the empty seats and the idle hours. It works less well as fares fall, and it fails where incomes make a high fare unaffordable, which is most of the market in most African cities.
Scheduling attacks both terms at once. If a passenger commits to a departure time and a route in advance, the operator can fill the car before it moves and can plan the driver's day as a sequence of loaded trips rather than a series of responses. Occupancy rises from one to three or four. Idle time falls. The cost per passenger falls by something close to the ratio, which is roughly what a halving of the fare implies.
The passenger pays for that with spontaneity. They must decide the night before rather than on the pavement.
That is the actual trade, and it is worth naming plainly: instantaneous matching is a premium feature, and most urban transport systems in the world have never offered it. A bus does not come when you summon it. Neither does a train, a minibus taxi on a fixed route, or a shared taxi that departs when full.
We argued recently that the robotaxi model developed in Austin is unlikely to transfer to Nairobi, Jakarta or Lagos, because autonomy substitutes capital for labour and its business case is strongest where labour is most expensive and capital cheapest. That is the negative version of the argument.
This is the positive version. The constraint in an African city is not that the vehicles are the wrong kind. It is that the vehicles are empty too much of the time and the fares are too high for the people who need them. Those are utilisation and affordability problems, and they are addressable with scheduling, aggregation and routing rather than with autonomy or electrification.
Which suggests a sequence that is the reverse of the usual one. Raise utilisation first, because that is what makes a vehicle affordable to operate and a fare affordable to pay. Electrify second, because high utilisation is precisely what makes an electric vehicle's economics work, given that its advantage is a lower cost per kilometre against a higher cost per vehicle. A vehicle doing 200 kilometres a day with four passengers is a far better candidate for electrification than one doing 60 with one.
So an operator that has solved utilisation has built the conditions for electrification without mentioning it. An operator that electrifies first, at low utilisation, has bought an expensive asset that is idle.
Density of demand along corridors, not across the city. Scheduled aggregation requires enough people going from roughly here to roughly there at roughly the same time. That is a property of a corridor, not of a population. A city can be enormous and still not support the model outside a handful of routes, and the failure mode is a service that works beautifully on three corridors and cannot expand to a fourth.
The reliability of the schedule. The passenger has surrendered flexibility and received a promise in exchange. If the vehicle is late twice, the promise is worth nothing and they return to whatever is summonable. Punctuality is not a service quality metric here, it is the product.
What happens when the car is not full. Three passengers instead of four is a twenty-five per cent revenue reduction on a trip whose costs did not change. Whether the operator or the passenger absorbs that, and whether a trip runs at all below a threshold, is the term that determines the unit economics and it is invisible from outside.
Load factor across the whole week, not at peak. The average number of paying passengers per vehicle-hour, including the return leg and the midday trough, is the number the whole business rests on and it is not what gets quoted.
What the passenger gave up. Whether a scheduled service reaches people who previously could not afford a car trip at all, or mostly converts existing ride-hailing users to a cheaper option. Those are very different outcomes, and only the first is an access gain.
Who is excluded by the schedule. Shift workers, traders with unpredictable days, people caring for others, and anybody whose life does not permit a commitment made the night before. A service organised around predictability selects for people with predictable lives.
Driver earnings per hour, not per trip. Higher utilisation can raise a driver's income or simply raise the operator's margin. Which of those happens is a contractual question, and it is the same question about who captures the gain that runs through every mobility transition.
The vehicle is the least interesting part of most mobility transitions. The operating model is where the cost sits, and it is the part that can change without anybody buying anything new.
The Lab works on this in e-mobility and transport, and our fieldwork with commercial riders in Nairobi is built around the same arithmetic of hours, fares and utilisation.
If you are assessing a mobility model in a city where the fare is the binding constraint, tell us what you need to know.
This is an independent insight piece by Transitions Lab. For the Lab's applied work, see E-Mobility & Transport. See also A Thousand Cars, One Risk on why the autonomous model is hardest to transfer where mobility need is greatest, and Own the Battery, Rent the Shopfront on how operating architecture rather than technology decides a network. To discuss a study, see Contact.