Insight, E-Mobility
Grab is putting a charging network inside its driver app. Yulu is renting out 200,000 electric bikes. Both are solving an infrastructure-financing problem by owning the demand. That works, and it concentrates three dependencies on one counterparty.
Charging infrastructure has a circular problem that everybody in the sector can recite. Chargers are not built because there are too few electric vehicles. Vehicles are not bought because there are too few chargers. The conventional answers are public subsidy, a mandate, or patience.
Two announcements this month describe a third answer, and it is the one that is actually working in Asia.
Grab has invested in the Vietnamese charging startup EBOOST, which the two companies describe as running around 2,500 charging points and serving upwards of 10,000 users at the time of the announcement. The network is being built directly into the Grab Driver app, so a driver finds a charger, starts a session, and pays without leaving the interface they already use for work. Company statements target more than 6,000 charging ports by early 2028.
Separately, India's Yulu has raised roughly US$93m to grow its electric two-wheeler fleet from roughly 50,000 vehicles to around 200,000 over two years, per its own announcements. Riders do not buy the vehicle. They subscribe. The company reports the fleet supporting hundreds of thousands of deliveries a day, mostly for quick commerce and food delivery.
Neither company is primarily an energy business. Both are doing something that energy businesses have never been able to do on their own, which is to make charging infrastructure bankable. The mechanism has a name, borrowed from a much older industry.
Shopping centres are not financed on the strength of the small units. They are financed on the department store that signs a 20-year lease before the ground is broken. Ports are financed on a shipping line's volume commitment. Telecoms towers are financed on a master lease with a mobile network operator. In each case the anchor tenant does not pay for the asset. It does something more valuable: it converts speculative demand into contracted demand, which converts an equity risk into a debt risk, which halves the cost of capital.
The binding constraint on charging infrastructure has never really been the price of the hardware. It is that nobody will lend against an asset whose revenue depends on a diffuse population of consumers who might show up. Utilisation is the whole business. A charger at 15% utilisation is a write-off and the same charger at 45% is an annuity, and the difference is not technical.
A ride-hailing platform with a concentrated driver fleet has something no utility and no charge-point operator has ever had: it knows where the demand is, when it will arrive, and how long it will stay, because it is dispatching it. Gig fleets also have duty cycles that are close to ideal for charger economics. High daily mileage, predictable dwell windows between shifts, and drivers whose income depends on not queueing.
Putting the charging network inside the driver app is the contractual instrument. It is a soft exclusivity, achieved through convenience rather than through a signed volume commitment, and it produces most of the same financing effect.
Yulu is running the same logic one layer down. A subscription fleet takes the asset off the rider's balance sheet, where it could never be financed at a sane rate, and puts it onto a corporate balance sheet where it can be. It also raises utilisation, because a vehicle used by a delivery worker for eight hours is worth several vehicles sitting outside a flat.
This is a real advance and it deserves to be recognised as one. It is also worth being precise about what has been built.
For the charging network, the anchor is the whole business case, and that is a position with two edges. The network's assets are sunk, immobile, and configured around one customer's duty cycle: sited near dispatch density, specified for that vehicle class, priced for that usage pattern. If the platform renegotiates, changes its in-app defaults, or builds its own network, there is no alternative demand to fall back on. The anchor tenant that made the asset financeable also holds an option on its future margin. Property developers have understood this about department stores for a century, which is why anchor leases run for decades and are lawyered heavily.
For the driver, the arithmetic is different and less discussed. In the fullest version of this model, one counterparty supplies the work, the vehicle, and the energy. Each of those relationships is individually beneficial and each is defensible on its own terms. Taken together they describe a person whose entire income-producing capacity is rented from a single firm, with no asset, no alternative dispatch, and now no alternative fuel.
This is not an accusation of bad intent. It is a description of a structure, and structures have predictable effects regardless of intent. The relevant precedent is not from technology. It is the historical pattern of employers who supplied housing, credit, and provisioning alongside wages: not because anybody set out to trap anyone, but because bundling was efficient, and because the accumulation of dependencies quietly changed what a worker could refuse.
Our argument in Who Absorbs the Gap was that in markets with unreliable power, adoption is decided by which party absorbs the volatility, and that moving the burden off the rider is usually the right move. That still holds. What the anchor-tenant model adds is that the party which absorbs a risk also acquires leverage from having absorbed it, and nobody is currently measuring what happens to that leverage over time.
There is a great deal of data on these fleets, and it is all held by the platform, and it all measures the platform's questions. Utilisation, session duration, charger uptime, driver retention, orders per hour. Useful operational metrics, none of which answer the question a funder or a regulator should be asking.
Four things are worth measuring independently, and none of them can be measured from platform telemetry.
Gross fares are not income. Subscription fee, energy cost, penalties, and idle time are all deductions, and all of them are set by the same party that sets the fare. The number that matters is what remains, measured at 12 and 24 months after enrolment rather than at signup.
This is the empirical version of the concern, and it is testable. Take rates, subscription pricing, and per-kWh charges can be tracked against the share of a driver's alternatives that remain available. If the terms are stable as dependency deepens, the concern is unfounded and the sector should be able to say so with evidence.
What proportion of the network's throughput comes from one counterparty, and what contractual term protects it. A charging network with 90% single-platform throughput and no volume commitment is not an infrastructure asset. It is a receivable.
Platform fleets electrify first because their duty cycles genuinely suit the economics. They also electrify first because platforms subsidise the transition to lock in supply. These produce identical adoption curves and completely different futures, and telling them apart requires knowing what the driver pays when the incentive ends.
The anchor-tenant model is the most effective mechanism currently available for getting charging infrastructure built in Southeast and South Asian cities, and it is going to be replicated. That is a good outcome. It is also a governance question arriving early enough that it can still be answered cheaply.
The uncomfortable version, stated plainly: this model transfers ownership risk away from the least capitalised participant, which is the right thing to do, and in the same movement transfers bargaining power away from them too. Both effects are real. Only one is being reported.
An independent baseline taken now, before 200,000 vehicles and 6,000 ports are in the ground, costs a fraction of what it costs to establish the same facts retrospectively during a dispute. That is ordinary field research and ordinary impact measurement, applied to a question the sector has not yet decided to ask.
The Lab works on this in e-mobility and transport, at its intersection with financial inclusion, and from direct fieldwork with commercial riders in Nairobi whose economics look a great deal like these. Whether a niche succeeds also depends on what the surrounding institutions do about it, which is the argument in Four Ways a Transition Lands.
If you are investing in a platform-anchored fleet or charging network and want the driver side measured by somebody who is not the platform, 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 and Financial Inclusion & Payment Systems. To discuss a study, see Contact. See also Who Holds the Pen on the Standard on the same asymmetry playing out in Kenya's open battery-swap network, The Smelter Contract on the same bargain scaled up to a national grid and a state government, and A Thousand Cars, One Risk on the same counterparty-dependence problem sharpened when the asset cannot be driven without the manufacturer's software.