Battery-as-a-Service and the New Infrastructure Race

Every new energy infrastructure arrives promising freedom from the last one.
When natural gas displaced coal for urban heating, the commentary at the time celebrated the democratization of energy — distributed, clean, no longer hostage to the coal merchant or the rail yard. What actually happened was that the people who owned the pipe networks did not lose their structural position. They tightened it, because pipe is harder to replicate than coal, and because consumers who have built their daily routines around one infrastructure rarely have a real alternative once the network reaches density. The technology changed. The control point did not.
Battery-as-a-Service, or BaaS, is being described in the same language every new infrastructure format attracts: cleaner, more flexible, more democratic, less captive. My read is different. BaaS is not eliminating the point of control; it is relocating it from the petrol forecourt to the swap station. And the operators who understand that earliest are not merely selling batteries in a new form. They are reconfiguring choice itself.

The Mechanism Is Visible in Nairobi

The clearest place to see this is not Beijing or Brussels, but Nairobi.

A boda boda rider is a useful lens because his problem is not abstract. He earns trip by trip, his fuel cost is volatile, his access to credit is thin, and his margin for error is small. In Choice Competition terms, he is not choosing from a stable category; he is choosing under constraint, with Cost, Access, Effort, Fit, and Perceived Risk all pulling at once. He needs energy access that is near, predictable, and financeable. He also needs the burden of ownership to disappear from his daily decision-making. That is a very specific resolution mode. It is not the same as the one a suburban commuter or a private car owner face.

This is why a battery-swap subscription can look, from the outside, like a financing tweak and, from the inside, function like an infrastructural capture. The rider is not just buying power. He is buying removal of uncertainty. The daily fee matters less than the fact that it converts a volatile operating input into a known obligation, timed to the rhythm of his earnings. That shifts the Cost dimension from unstable to legible, lowers Effort at the moment of use, reduces Perceived Risk around battery ownership, and makes Access something the operator controls rather than the rider negotiates each day. The position changes across multiple dimensions at once. That is the part people miss.

M-KOPA is the best illustration I have seen because it understands that the battery is the acquisition mechanism and the relationship is the asset. It has financed over 5,000 electric motorbikes for boda boda riders, and those riders report meaningful daily savings and higher earnings. More importantly, M-KOPA has built credit history, income-smoothing, and equipment-financing tools onto the repayment relationship, so the rider who starts with a battery is gradually pulled into a broader financial infrastructure. The swap station is visible. The dependency is not. It sits behind it, like the feeder canal behind an irrigation gate.

That is not a footnote. It is the business.

The strongest objection is straightforward, and I think it deserves to be stated fairly. Battery packs are getting cheaper, and learning curves are real. If the underlying component falls far enough in cost, ownership should win back the field, just as residential solar leasing weakened once panels got cheap enough. On that view, BaaS is a bridge. It exists while batteries are expensive, and then it fades. That case is coherent. It has the virtue of using a real historical pattern rather than a slogan.

But the analogy breaks where the market becomes a network rather than a component. A cheaper battery helps the consumer who wants to own outright. It does not help the operator who has already built density. A denser network compounds with each added station, because reliability improves, usage rises, placement gets better, and the data generated by actual swaps sharpens the next deployment. That is not a cost curve. It is positional hardening. One is a factory input that gets cheaper with scale. The other is a port whose value rises because more ships keep coming through it. They are not the same economic object.

This is why I think Nairobi matters. It strips the mechanism to its essentials. The rider is not comparing two batteries. He is comparing two operating positions. One position leaves him exposed to volatility and ownership risk. The other wraps energy access, payment timing, and credit formation into a single relationship. Once that relationship exists, the comparison no longer runs primarily through battery chemistry. It runs through dependency. And dependency, once it has enough daily repetition behind it, becomes very hard to dislodge with a better product.

I could be wrong about how quickly that hardening arrives in each geography. The path in East Africa will not be the same as the path in China, and neither will look exactly like India or Europe. But the underlying mechanism seems clear to me. BaaS is not merely changing the economics of fuel. It is changing the competitive distance between the operator and the user.

That is the real shift.

The Consensus Has the Right Mechanics and the Wrong Variable

The strongest version of the opposing view does not deny that BaaS operators are building relationships. It denies that those relationships are durable enough to matter structurally. And it has the better of the historical evidence — at least on the surface.

The argument runs like this. BaaS is a financing innovation, not a structural one. It exists because battery costs are still high enough to make outright ownership prohibitive for price-sensitive consumers — the boda boda rider, the fleet operator, the urban commuter choosing an entry-level electric vehicle in a market where the upfront cost difference between an ICE and an EV is still significant. Remove that cost barrier, and the logic that makes a subscription attractive unravels. The rider who once needed M-KOPA to absorb the battery's capital cost can, in a world of cheap enough packs, simply own the asset. The dependency dissolves. The operator loses the structural lever.