First edition · August 2026

Market insights report

Understanding Nigeria’s EV Ecosystem

Observed demand, rider economics and infrastructure performance across Lagos and Abuja, with an outlook to 2030.

56,973vehicles counted

39charging and swap sites audited

106drivers interviewed

12corridors, two cities

leke.services

—

Introduction

Why this document exists

Most of what is written about electric mobility in Nigeria is either an announcement or a forecast. This is the third thing: a record of what was actually there.

Nigeria’s EV story is moving faster than most people realise. In only a few years we have gone from a handful of imported vehicles to locally assembled EVs, growing charging networks and real policy momentum. This first edition pulls that story together in one place: what is happening, who is building it, and where the road leads.

It is written for anyone with a stake in electric mobility in Nigeria — policymakers, investors, operators and curious readers alike. Every EV on Nigerian roads means cleaner air in our cities, quieter streets and a real dent in emissions, while getting more people access to affordable transport.

·About Leke Services

Leke Services is a management consulting firm that helps leaders navigate the mobility energy transition. We help fleet operators, charging companies, financiers and manufacturers make practical decisions by understanding market opportunities, refining operations and handling stakeholders. We develop initiatives from idea to execution, and operate in both Nigeria and the United States. This report grew out of that work, and out of the questions our clients keep asking.

·What we did

Between February and March 2026 our team stood on twelve corridors in Lagos and Abuja and counted vehicles. We visited 39 charging and battery-swap sites to see whether they were open. We asked 106 drivers what they earn and what they spend. We interviewed senior people across vehicles, infrastructure, fleets and finance about what they are seeing.

Announcements tell you what somebody intends to build. Forecasts tell you what a spreadsheet believes. We wanted the third thing: a record of what was actually there on a Tuesday afternoon on Ikorodu Road.

It is a partial record, and we have tried to be precise about its edges. Twelve corridors is not a country. One hundred and six drivers is not a population. A single visit to a charging site tells you whether it was open that afternoon, not whether it is reliable. Chapters one to seven describe what we observed. Chapter eight is different in kind: it is a model, and it states the assumptions it is built on.

12 corridors counted

56,973 vehicles observed across three time windows in Lagos and Abuja, of which 961 were electric.

39 sites audited

Charging and battery-swap sites visited once each, recording status, utilisation and energy source.

106 drivers interviewed

Intercepted at swap stations, charge points and ranks, and asked what they earn and what they spend.

·Who this is written for

·How to read the evidence in this report

Readers who want the whole argument in five pages will get away with reading only chapter one. Readers who intend to deploy capital should read chapter one, then the chapter matching the question above, then the relevant appendix, because that is where the underlying data sits.

Every substantive claim carries one of four grades. They describe how far the evidence reaches, not how confident we feel. Where a claim is our judgement rather than an observation, we say so in the sentence.

·A note on operator names

Utilisation and availability figures here come from single-visit observations across 39 sites. That is enough to describe how the market performed in aggregate. It is not enough to grade an individual company, and publishing it under brand names would invite readers to treat a snapshot as a verdict on a business.

Operator performance is therefore reported against letter codes that carry no ordinal meaning, with site counts given as bands. Named operators appear only in the site register at appendix B.6, which records presence and location and contains no performance data. We have kept the two apart deliberately. With roughly a dozen operators active and network sizes broadly known in the sector, an informed reader may still draw inferences. What anonymisation removes is this report’s editorial judgement on any named company.

·With gratitude

• Sterling Bank — for sponsoring this first edition.
• Qoray Mobility & Energies — for their transparency and openness, and for supplying operating data that almost no operator in this market publishes.
• The Leke Services team — for the diligent work behind every page of this document.
• Joshua Popoola and his team — for the field work that grounds this document in reality.
• NADDC — for strengthening Nigeria’s manufacturing capability and ushering in a more efficient mobility future, one that includes EVs and well-thought-out infrastructure.
• The 106 drivers who stopped mid-shift to answer questions from strangers with clipboards, and who were, without exception, better informed about their own economics than most of the analysis written about them assumes.

01

Chapter one

Executive summary

Across twelve corridors in Lagos and Abuja we counted 56,973 vehicles, of which 961 were electric. That is 1.69% of the vehicles we saw, against a modelled 0.4% of the national vehicle parc. Both measure something real. Most of what follows is an attempt to read the gap between them carefully.

1.1What we did

Field counts are the spine of this analysis. Where we model forward, we say so and show the assumption. Where we saw something once, we call it observed and do not dress it as a trend.

1.2Ten figures from the field work

Each figure states what was measured and the base it was measured on. Denominators differ between them, so they should not be combined.

1.3Five observations

The two most-quoted penetration figures for Nigeria measure different things and are routinely confused. EV share of the national vehicle parc is 0.4%, a stock measure. EV share of vehicles passing our observation points is 1.69%, a flow measure. On the corridors sampled, electric vehicles were 4.2 times more visible in traffic than the modelled fleet share would imply.

A gap of that size is unlikely to be a rounding artefact. The most straightforward reading is that the electric vehicles operating in these corridors are largely working assets on commercial duty cycles rather than private cars parked for most of the day. A delivery rider covering 100 km passes an observation point many times; a privately owned saloon does not. If that reading holds, the flow measure is the more relevant one for sizing a charging or swap business, because revenue follows kilometres rather than registrations.

Lagos corridors returned 2.38% observed penetration against 0.33% in Abuja, a seven-fold gap. Composition diverged further. In Lagos, 53% of electric vehicles observed were two-wheelers of the type used in logistics and delivery, in corridors served by a dense battery-swap network. In Abuja, 95% were four-wheelers, served by a handful of charging points. Of 64 electric vehicles counted in Abuja, three were two-wheelers and none were three-wheelers.

We did not survey enough Abuja operators to explain the divergence with confidence. What the counts suggest is that a swap-led model built around Lagos two-wheeler fleets would find little comparable demand in the Abuja corridors we observed, and that the reverse also appears to hold.

The 106 drivers we intercepted reported average gross revenue of ₦20,777 per operating day. Energy, whether swap or charge, cost ₦4,085, or 19.7% of revenue. That sits below 25% of gross revenue, the level we treat as the point at which commercial vehicle operation stops paying. After a lease payment of ₦5,000 to ₦7,000, riders retained ₦9,700 to ₦11,700 a day.

That is a workable margin rather than a generous one, and it is reported by drivers rather than modelled by us. 96% of those surveyed reported cost savings against petrol equivalents. Within this sample, which is Lagos-heavy and two-wheeler dominant, demand-side viability appears established. We would not extend the finding to four-wheelers or to Abuja on the evidence collected.

Of 53 visits across 39 unique sites, 37 found the site operational. That headline masks the problem. Every battery-swap site visited in Abuja, across two separate networks, was inactive, with nearby stakeholders reporting downtime of several weeks to months. Estimated capacity utilisation ranged from 55% at the largest network down to 5% across seven smaller operators. Grid-dependent sites underperformed hybrid solar-and-diesel sites consistently and without exception.

Across the sites audited, the constraint that presented was not the number of charge points but whether a given point was open, discoverable and powered at the moment a rider arrived. That reads as an operating and energy-supply problem rather than a capital-availability one, though a single-visit audit cannot rule out timing effects at individual sites.

Just 7.5% of drivers surveyed owned their vehicle outright. Lease-to-own accounted for 48.1% and company ownership for a further 34.0%, so 82% of adoption in this sample was intermediated by a fleet or a financier rather than bought directly. On that basis, adoption in the segments observed looks closer to a balance-sheet decision than a consumer-purchase one.

If that holds more widely, the cost of capital matters more than the sticker price. At the modelled 2026 EV fleet lending rate of 35%, four-wheeler fleet economics do not clear. Our view, and it is a view rather than an observation, is that cheaper credit for fleet operators would move this market further than a subsidy on vehicles.

1.4What this may imply if you are building

1.5What would change our view

• Battery degradation data. Half of drivers surveyed named it their leading concern and no operator we spoke to could produce state-of-health data across a cohort. If real-world degradation proves worse than assumed, lease-to-own residuals break and the financing model with them.
• A tariff shift. Charging economics rest on a spread between grid cost and retail rate. A Band A move in either direction changes the site-level return materially.
• Enforcement of local content. The 30% local-content target sits in a bill, not a law. If it is enforced, import-led strategies reprice. If it is not, local assembly investment does.
• Cost of capital. The model has EV fleet lending falling only from 35% to 30.5% by 2030. Nothing else in this report has the leverage that a faster fall would.

02

Chapter two

Where the demand is

Three of the corridors we observed returned penetration above 3% and three returned below 0.3%. Four of those six sit within twenty kilometres of each other. On this evidence, where the first hundred assets are sited appears to matter more than how many are bought.

2.1Flow against stock

Nigeria’s electric vehicle penetration is quoted at figures ranging from below 0.3% to nearly 2% depending on the source. Both ends of that range appear in this report. They are not in conflict, and understanding why is the first step to sizing anything in this market.

A stock measure asks what share of registered or estimated vehicles are electric. On the model’s Base case, 49,728 electric vehicles against a parc of 12.3 million gives 0.4% for 2026; against the wider 20 million parc used in earlier work, 0.25%. A flow measure asks what share of vehicles passing a point over a given window are electric. Our corridor counts give 1.69%.

The multiple between them is a direct measure of how much harder Nigeria’s electric vehicles work than the average vehicle on the road. It is the most useful number in this report for anyone deploying infrastructure, because charge points earn from throughput, not from registrations. It also means headline national penetration figures systematically understate the addressable market for energy services and systematically overstate it for anything sold to private owners.

2.2The corridor picture

Twelve corridors were observed, eight in Lagos and four in Abuja. Penetration ranged from 4.83% on the Maryland to Ikorodu Road stretch down to 0.13% on Apapa Wharf Road, a thirty-seven-fold spread within a single sample. Six corridors sat above the 1.69% sample average and all six are in Lagos.

The corridors above the sample average shared three observable characteristics: dense logistics and delivery activity, proximity to a battery-swap site, and commercial fleets operating on them. Maryland to Ikorodu Road, the Lekki–Ikoyi Link and Ozumba Mbadiwe all sit within the Lagos delivery economy. Apapa Wharf Road, the access corridor to Nigeria’s largest port complex, returned two electric vehicles out of 1,553, consistent with the absence of an electric heavy-freight option in this market. We did not test these associations statistically and present them as observed co-occurrence, not established causation.

Corridor detail

Shaded rows are corridors above 3% observed share. Source: Leke Services corridor observations, March 2026.

2.3Two cities, two markets

Aggregating Lagos and Abuja into a single national figure obscures the clearest pattern in the dataset. Lagos returned 897 electric vehicles across 37,703 observed, a 2.38% share. Abuja returned 64 across 19,270, a 0.33% share. The seven-fold gap in level is notable; the difference in composition is arguably more so.

2.4When the vehicles are moving

Penetration was highest in the evening peak at 1.89%, marginally ahead of the morning peak at 1.86%, and lowest across the long midday window at 1.52%. The spread is narrow and we would not read much into the gap between the two peaks. The direction is at least consistent with commercial fleet operation, where logistics returns and ride-hailing demand coincide in the evening.

For infrastructure operators the practical reading concerns queue design rather than total demand. More than half of all vehicles observed passed during the midday window, when electric share was lowest. Capacity sized to average daily throughput would be expected to run short at the evening peak and idle through much of the afternoon, a profile broadly consistent with the utilisation figures in chapter four.

2.5What this implies for siting

• Traffic volume was a poor predictor of EV count. The three highest-penetration corridors were logistics arteries rather than the busiest roads. Garki Area 1 in Abuja carried the highest absolute vehicle count in the sample, 9,829, and returned fifteen electric vehicles.
• Every above-average corridor sat within the existing Lagos swap footprint. No corridor we observed without nearby swap infrastructure exceeded 1.3%. Whether infrastructure follows demand or creates it is not something this dataset can settle.
• Demand shape varied more than demand level. The narrow peak-to-trough spread in penetration sits against a wide spread in absolute volume, which points to time-of-use pricing as a lever that appears little used in this market.
• The Abuja corridors did not resemble the Lagos ones. Different dominant segment, different infrastructure form. On this evidence they warrant separate assessment.

03

Chapter three

Who is already here

Twelve operators run charging or battery-swap infrastructure across the sites we audited. Between them they hold roughly 39 sites we could find and verify. That is a small industry, and its shape tells you more about where the openings are than any market-size figure.

3.1The shape of the operator base

The operators we audited fall into three groups that behave differently enough to be worth separating.

The third group is the one most often miscounted in market maps. A closed charging network attached to an owned fleet is a cost centre that works, not a charging business. Counting its sites as public infrastructure overstates the network available to an independent rider by a material margin.

3.2Business models observed

Four models were visible in the field. They are not mutually exclusive and several operators run two.

• Battery swap for two- and three-wheelers. The dominant model by vehicle count. Swap at the bay in two to three minutes, energy sold per swap rather than per kWh, and the battery stays on the operator’s balance sheet. This is what makes the rider economics in chapter five work: the rider never buys a battery.
• Destination charging for four-wheelers. Sites at hotels, malls and offices. Generally within a host’s premises, which constrains signage and access hours. Utilisation across the four-wheeler charging operators audited ran from 5% to 28%.
• Fleet-captive charging. Depot infrastructure serving an owned fleet. Reliable because it has to be, but not accessible to anyone else.
• Hybrid installation and operation. Operators who build infrastructure for others as well as running their own. A small group, and the one most likely to consolidate if the market grows.

3.3What the operator base is missing

Setting the audited operators side by side, three gaps are visible from the data rather than inferred from opinion.

3.4Case study: full value-chain participation

3.4.1 The coordination failure it responds to

Nigeria’s EV market suffers a coordination failure. Vehicles cannot scale without charging. Charging cannot justify investment without vehicle volumes. Neither can grow without financing, and financing will not arrive until the first two are demonstrable. Every participant is waiting on someone else to move first.

Qoray operates meaningfully across all four value pools at once. Operating in four pools at once could be read as a lack of focus. We read it as a response to the coordination failure: an integrated operator can resolve internally what the market cannot yet resolve between firms.

3.4.2 The operating record

The figures below are Qoray’s own, covering the first quarter of 2026 and supplied for this report. They are not observations of ours and we have not audited them. They matter because almost no operator in this market publishes anything comparable, which is the disclosure gap chapter five identifies as the binding constraint on lending.

The Kano operation is the most interesting line in the set, and it sits outside the geography of our own field study. Electric tricycles on battery swap covered 150,000 km across almost 4,000 swaps, with many drivers swapping more than once a day. That multiple-swap behaviour is the clearest available evidence that swap infrastructure removes the multi-hour charging downtime which otherwise caps commercial utilisation. It also indicates that the swap model travels beyond Lagos, which nothing in our corridor data could have told us.

3.4.3 Why the operating data matters to a lender

Operational intensity is functioning here as financial infrastructure. Granular telematics covering daily kilometres, energy consumption, driver behaviour and revenue per kWh turns a vehicle from an opaque depreciating asset into a transparent, monitorable one that a lender can underwrite.

• Data-driven procurement. Real-world mileage data determines each vehicle’s range and battery specification before import, which removes speculative inventory risk from the balance sheet.
• Driver behaviour as an asset lever. A battery rated for 500 km delivers barely 300 km without disciplined driver behaviour, according to one operator’s telematics. Treating human factors as a core asset-management variable, rather than a training afterthought, is what makes the range assumption in a credit file hold.
• The next frontier is after-sales. Diagnostics, parts supply and battery second-life are, on this operator’s reading, the most under-appreciated value pool as the parc scales to tens of thousands. That matches what our driver survey found about technician access.

That statement is the single most consequential sentence a lender has said on the record about this market. It is a bank confirming that an EV book has performed at least as well as a conventional auto book. Chapter five identifies seasoned repayment data as the first thing lenders said they needed; this is the first instance of it being described publicly.

3.4.4 What this suggests, and what it does not

The reading we take from the case is narrower than the one usually drawn from it. It is a claim about sequencing, not necessarily about superiority.

04

Chapter four

Infrastructure reality

Thirty-seven of 53 site visits found the site open. That sounds like a functioning network. Underneath it, utilisation ran from 5% to 55%, one city’s entire swap footprint was dark, and the difference between the operators that worked and the ones that did not was, in every case, how they got their power.

4.1Coverage

The audited sites concentrate heavily in Lagos, and within Lagos in the corridors identified in chapter two. Abuja carried a nominal footprint that was, at the time of visit, largely non-functioning. We do not treat 39 sites as the total network; it is the number we could locate, reach and verify.

4.2Utilisation and availability

Estimated capacity utilisation across the twelve audited operators spanned an order of magnitude. Two swap networks ran near half their estimated capacity. Seven operators sat at roughly 5%. One four-wheeler charging operator was at 28% utilisation but only 33% site availability, a combination that suggests demand exists at the sites that are open.

The pattern that held without exception was the energy one. Every operator running hybrid, solar or self-generated power held higher availability than every operator running predominantly on grid. The largest network in the sample, and the only one with more than twenty audited sites, was also the only one recorded with low energy reliability.

4.3The visibility problem

Several sites were physically present but effectively invisible: no external signage, restricted access through a host’s premises, and operating hours published nowhere a rider would find them. In at least one operator’s case the network is designed for fleet and opportunistic use rather than open public access, which is a legitimate strategy. The consequence is the same for the market: capacity that exists but cannot be planned around does not function as capacity.

4.4What the sites suggest about siting

• Sites inside a host’s premises inherit the host’s hours and the host’s signage rules, and both suppress utilisation.
• Grid-only sites in the audited set did not sustain availability. Every operator that held 100% availability had some form of self-generation.
• The operators with the highest observed utilisation were those with fleet offtake attached, not those with the best locations.
• No audited operator published operating hours in a form a rider could check before travelling.

05

Chapter five

Does it pay?

Start with a rider. Drivers surveyed reported gross revenue of ₦20,777 a day, of which roughly a fifth went on energy, leaving ₦9,700 to ₦11,700 after the lease. Those are working numbers at the rider level. Whether they yet constitute a financeable asset class is a separate question.

5.1Where the money goes

One hundred and six drivers were intercepted across Lagos and Abuja and asked what they earn, what they pay for energy, and what they pay for the vehicle. The answers are the most granular unit economics available for this market, and they are better than most observers assume. They are also self-reported, and we did not verify them.

Energy at 19.7% of gross revenue sits below 25%, the level we treat as the point at which commercial vehicle operation stops paying. Riders we spoke to who had previously operated petrol motorcycles described energy costs well above that level, and our model has petrol at ₦1,350 per litre in 2026 rising to ₦1,900 by 2030. That is consistent with the 96% reporting savings against their previous vehicle, though we did not verify the comparison independently.

5.2The margin band, and what moves it

Two things follow. First, for the drivers surveyed the lease was a larger cost line than energy. Second, the energy line is not a pass-through of the electricity price: the model marks charging operators up 1.30x over grid cost, implying that a little under a quarter of a rider’s energy spend is funding the operator’s margin as the power networks are built out.

5.3Who actually owns these vehicles

Only 7.5% of drivers surveyed owned their vehicle outright. Lease-to-own at 48.1% and company ownership at 34.0% together account for 82% of the sample, with a further 10.4% on daily rental. In the segments observed, the purchase decision sits with an institution rather than with the rider operating the vehicle.

Consistent with that, 45.3% of drivers reported technician access only through their company, against 48.1% with direct access. In this sample maintenance capability sits inside fleets rather than in an open market. For an independent rider, the absence of a workshop they can walk into would be a material barrier, and several stakeholders raised it unprompted.

5.4What drivers say is stopping them

Asked whether they would adopt or continue with an electric vehicle, 46.2% said yes without qualification, 48.1% said they would only if conditions improve, and 5.7% said no. The two positive answers are often reported together as 94%. We show them separately, because a conditional yes is a different commercial fact from an unconditional one: it tells us the driver is willing, not that the driver will act.

The conditions cited were specific and consistent. Battery degradation led at 50.5%, followed by limited charging infrastructure at 44.7% and unreliable electricity at 35.0%. Charging time and range followed. Purchase price did not appear in the top five, which is unsurprising in a sample where 82% of riders never see it, and should not be read as evidence that price is not a barrier in the wider market.

5.5What lenders said they still need

Rider economics and stated willingness both came back positive in this sample. What lenders described as missing was the evidence base needed to price the risk.

On the evidence collected, most of these read as disclosure gaps before they are capital gaps. The operators holding the data that lenders asked for are largely the same operators who would benefit from it being known. Whether they have a commercial reason to withhold it is a question we did not put to them directly, and it is one we would want to answer before drawing a firmer conclusion.

06

Chapter six

What is blocking scale

Every stakeholder we interviewed named the cost of capital before they named anything else. The field data supports them: the lease is a bigger line in a rider’s day than the electricity, and the lease is priced off a lending rate our modelling does not project to fall much before 2030.

6.1The cost of capital

The model carries an EV fleet lending rate of 35% in 2026, easing to 30.5% by 2030 as the policy rate falls from 26.5% to 22%. Prime lending sits between the two. Against those levels, the operators we interviewed were consistent about what they need, and it is not what the curve delivers.

The gap is not marginal. On the model’s own arithmetic, four-wheeler fleets do not clear at the lending rates available, and two-wheeler fleets clear because the asset is cheap and the duty cycle is intense, not because the financing is good. That asymmetry explains most of the segment mix in chapter two.

6.2The other four constraints

The foreign exchange constraint deserves a note of caution. A currency path to 2030 is among the least reliable things any model produces, and this one is no exception. What is more defensible than the level is the direction of the exposure: with 97.5% of electric vehicles imported at the time of the field study, this market prices its assets in a currency its operators do not earn.

6.3The skills constraint has a safety dimension

The technician shortage described above is usually framed as a capacity problem: not enough trained people to service a growing fleet. Stakeholders we interviewed raised a second dimension unprompted, and it is also serious. High-voltage systems are being worked on by people without high-voltage training or equipment.

We did not conduct a safety audit and are not in a position to quantify this. What we can report is that no operator we spoke to could produce a documented high-voltage safety protocol, and that 45.3% of drivers reach a technician only through their employer, which means the training standard applied to a vehicle depends entirely on which fleet happens to own it.

This matters commercially as well as morally. An insurer pricing EV fleet cover, or a lender underwriting a battery as collateral, is exposed to a workshop practice neither of them can currently observe. Our view is that a certification standard for high-voltage work is a prerequisite for the insurance products this market says it wants, not a separate social good to be addressed later.

07

Chapter seven

What policy allows and prevents

Nigeria has an electric vehicle policy agenda, a bill at second reading, a development fund in formation and a set of incentives that almost nobody is claiming. The gap between what is on paper and what is operating is the most consequential thing in this chapter.

7.1The instruments that exist

7.2The enforcement gap

Several themes recurred across the policy interviews and are reported here as stated rather than verified by us.

• Incentives are poorly communicated. Regulators themselves acknowledged that many market participants are unaware of existing tax holidays and CKD benefits.
• Customs data was not accessible. Import records on EVs, components and chargers were not available to us during the study, which limits any independent check on import volumes.
• Charging standards are fragmented across operators, with no unified interoperability requirement in force.
• Insurance and registration compliance was described as very low, including for coverage that is nominally mandatory.

The reading we take from this is narrower than the usual call for more policy. The instruments largely exist. What was missing, on the accounts given to us, was communication of them and capacity to enforce them. Those are different problems from the absence of a framework, and they are cheaper to fix.

7.3Standards, and why they matter commercially

Battery formats, connector types, payment rails and access control differ by operator across the sites we audited. For a rider this means being locked to one network. For a financier it means an asset whose usefulness depends on a single counterparty staying solvent. For a policymaker it is the cheapest available intervention, because a standard costs nothing to legislate and unlocks utilisation across every existing site.

08

Chapter eight

Outlook to 2030

This chapter is different from those before it. Chapters two to seven describe what we observed. What follows is modelled: every figure comes from the Base scenario of the Project Kiko Economic Model, the assumptions are listed at appendix F, and the sensitivities are stated where they matter most. Read it as a structured expectation, not as a finding. Figures in the companion financing paper are constructed differently: they are the unweighted average of six modelled world-states rather than the Base case alone, which is why its 2030 activity and capital figures are higher than those below. Both come from the same model.

8.1Vehicles

The Base scenario puts 462,693 electric vehicles on Nigerian roads by 2030, from 49,728 at the end of 2026. Two- and three-wheelers make up 93% of that parc. Annual new EV sales reach 190,752 units in 2030, against total vehicle sales of 2.35 million, an EV share of 8.1% of new sales. The total-sales base is drawn from the International Trade Administration’s Nigeria automotive sector guide.

The penetration assumptions behind it are segment-specific: two-wheelers reach 10% of new sales by 2030, three-wheelers 8%, four-wheelers 3%, buses 5%, light commercial 2% and freight 1%. The two-wheeler assumption is anchored to Kenya, where electric motorcycles reached 15.3% of new registrations in 2025 after an eight-year build. The Base case reaches 10% in five years against that 15.3% in eight.

Source: Project Kiko Economic Model, Base scenario. Totals may not sum exactly because of rounding.

8.2Infrastructure

Serving that parc requires 22,655 chargepoints by 2030, of which 9,062 are battery-swap stations for two- and three-wheelers and 7,929 are Level 2 AC points. The network is 2,441 points in 2026, so the build is roughly nine times the current base over four years.

Two caveats belong with that number. It is a requirement derived from vehicle-to-chargepoint ratios, not a forecast of what will be built. And the field evidence in chapter four suggests the market’s problem is not the count but the uptime, so a network of 22,655 points at the availability we observed would deliver materially less than the ratios assume.

8.3Energy

Annual energy demand reaches 1,149 GWh by 2030 from 112 GWh in 2026. Peak power demand is 289 MW. The generation mix shifts materially over the period: grid supply falls from 73% of delivered energy in 2026 to 50% by 2030, with standalone solar and storage rising to 35%.

That off-grid share is the most commercially interesting number in this chapter, and it is consistent with what we observed. The operators holding availability in the field were the ones with their own generation. The model assumes the market solves this by building generation alongside charging, at ₦165 billion of capital over the period.

8.4Capital required

Building the vehicle parc, the charging network and the generation behind it takes ₦2,456 billion between 2026 and 2030, roughly ₦2.5 trillion. Vehicles are 65% of it. Charging infrastructure is ₦543 billion, which is the figure most often quoted for this market and is a charging-only subset rather than the ecosystem total.

8.5Where the revenue sits

Annual revenue across the four value pools reaches ₦5,703 billion by 2030, from ₦394 billion in 2026. Fleet operations account for ₦4,749 billion of that, or 83%. Vehicles are ₦744 billion, enablers ₦164 billion and charging ₦46 billion.

The charging pool is the one worth pausing on. At ₦46 billion of annual revenue in 2030 against ₦543 billion of cumulative capital, charging is the most capital-hungry and least revenue-generative pool in the model. That is not an argument against building it. It is an argument that charging economics depend on offtake contracts with fleets rather than on walk-up demand, which is what chapter four’s utilisation figures already suggested.

09

Chapter nine

Where to play

The preceding chapters describe a market that works at the rider level, is constrained at the infrastructure level, and is financed almost entirely by institutions. This chapter sets out what that combination implies for anyone deciding where to put capital. It is our view rather than an observation, and it is labelled as such.

9.1Five entry points

9.2Sequencing

The order matters more than the selection. Three points follow from the evidence rather than from preference.

• Site against demand you can already see. Every corridor above the sample average sits within the existing Lagos swap footprint. Opening new geography ahead of demand was not something we saw working anywhere in the sample.
• Attach offtake before capital. The operators with the highest observed utilisation had fleet relationships. The ones at 5% did not. On a ₦12 million swap station that difference is the whole investment case.
• Solve power before scale. The largest network in the audit had the weakest energy reliability. Scaling a grid-dependent footprint appears to compound the problem rather than dilute it.

9.3What we would want to know before committing

If we were deploying our own capital into this market tomorrow, these are the four things we would pay to find out first, in this order.

• Real battery state-of-health across a cohort of at least 500 two-wheelers with more than eighteen months of service. Nothing else in the diligence stack matters if residuals are wrong.
• Site-level throughput data from two operators over a full month, to test whether our single-visit utilisation estimates hold up over time.
• Repayment performance on a lease-to-own book with at least one seasoned vintage.
• Whether Abuja’s four-wheeler demand is contractable — institutional, and able to underwrite a site — or private and diffuse.

A

Appendix A

Study overview and corridor observations

A.1Coverage

A.2Research approach

• Corridor observations. Vehicle counts and EV identification at fixed points, with time-of-day coverage across morning, midday and evening.
• Infrastructure audit. Charging and swap station visits recording operational status, utilisation and energy source.
• Driver and rider intercepts. Economics, usage patterns and adoption sentiment, captured at points of use.

A.3Limitations, as recorded by the field team

• Corridor observations vary slightly in frequency, so penetration rather than absolute count is the comparable metric.
• Analysis is based on sampled corridors, not full market coverage.
• Some infrastructure sites were inaccessible or non-operational at the time of visit.
• Driver responses are self-reported and were not independently verified.

A.4Corridor dataset

A.5Demand clusters, as tiered by the field team

A.6Distribution by time window and by state

Reconciliation: the three cuts above are independent tabulations of the same observations. Corridor totals, time-window totals and state totals each sum to 56,973 vehicles and 961 electric. No weighting or imputation has been applied.

B

Appendix B

Infrastructure audit and analysis

Twelve operators, 39 audited sites, one visit each. Operators are de-identified throughout. The site register at B.6 names operators but carries presence and location only, with no performance data attached.

B.1Operator comparison

Prices are as posted or reported at the site on the day of visit and are not a market survey. Estimated network scale is the operator’s own stated footprint where given, not a figure we verified. The audited column counts visits, not unique sites; several sites were visited more than once. See the open item at A.3.

B.2Strengths and gaps by archetype

B.3Energy mix and why it matters

Roughly 45% of the infrastructure audited relies solely on grid power. That is the single structural finding of the audit, because it exposes the network to systemic downtime in a market where grid supply is not dependable.

• Structural risk. Heavy reliance on grid-only power creates systemic vulnerability across the network rather than isolated failures at individual sites.
• Partial mitigation exists. Some operators deploy hybrid solar and diesel systems, but adoption is not widespread.
• Reliability leaders are differentiated by energy, not by scale. Operators with hybrid setups demonstrated stronger uptime and operational consistency in every case observed.

B.4Utilisation and reliability

• Access does not equal true reliability. The operator at 62.5% site availability also carried the weakest energy reliability, so a site being open does not mean a charged battery is available in it.
• Energy architecture drives performance. Hybrid-powered networks demonstrated stronger operational consistency across every comparison in the audit.
• Scale and reliability are fragmented. No operator combined high utilisation, high access reliability and high energy reliability.

B.5How utilisation and reliability were estimated

These are field estimates, not instrumented measurements. The method is set out here so a reader can judge how much weight to place on them.

B.6Site register

Presence and location only. Operators carry the same letter codes used throughout this appendix, so the register can be read alongside B.1 without naming any company against a performance figure. Site descriptors have been reduced to location and type; brand names have been removed from them.

Sites found operating at the time of visit

Sites recorded as not operating, inaccessible or not found at the time of visit

37 records in the first table and 20 in the second, 57 in total. Operators M and N appear in the register but were not part of the twelve-operator audited set, because no completed visit produced performance data for them. See the open item on site counts at A.3. Every battery-swap site visited in Abuja, across two separate networks, was inactive at the time of visit. Four-wheeler charging sites recorded as closed were visited at weekend afternoons and evenings; several operators run restricted hours that are not publicly posted, so a closed record is not evidence of a closed business. CHA-ABJ-06 is assigned to two different sites in the source records, operated by different companies; both are retained as recorded pending correction.

C

Appendix C

Driver economics and adoption

One hundred and six drivers intercepted at swap stations, charge points and ranks. Lagos 93, Abuja 13. Two-wheeler dominant. All figures are self-reported and none were independently verified.

C.1Unit economics

C.2Implied driver margin

Net take-home of ₦9,700 to ₦11,700 a day means drivers retain roughly half of gross revenue after energy and financing. The economics are profitable but not excess-margin. Earnings and costs vary materially across drivers, and the field team recorded four sources of that variation: vehicle type, route density between urban and peri-urban, the platform the driver works through, and access to reliable swap or charging.

C.3Cost perception

96.2% of drivers reported some level of saving against their previous vehicle. On this evidence EV adoption in the segments surveyed is economically rational and not subsidy-dependent at current scale.

C.4Adoption willingness

These are frequently reported as a single 94% figure by combining the first two responses. We report them separately throughout: a conditional yes is a different commercial fact from an unconditional one.

C.5Ownership structure and ecosystem support

About 82% of adoption in this sample is non-individual, being fleet-owned or financed.

C.6Adoption barriers

Primary structural barriers

Secondary economic and ecosystem barriers

Multiple responses were permitted, so shares exceed 100%. Financing and ecosystem gaps exist but sat well below infrastructure reliability in what drivers reported.

D

Appendix D

Field observations and on-ground realities

Two documented events, recorded as they happened. They are the clearest evidence in the study that this market’s constraint is energy before it is anything else.

D.1Case one: a demand shock

Swap site. Operated by Operator C, a mid-scale two-wheeler swap network. Observed at 09:00.

The site was open. It was staffed. It had batteries. None of that mattered, because the batteries were flat. This is the distinction between availability and reliability made concrete: an operational site with no charged stock is, from a rider’s point of view, a closed site with better lighting.

D.2Case two: a supply failure

Swap site. Operated by Operator G, the largest two-wheeler swap network in the audit.

The contrast between the two cases is the finding. At the first site a power failure produced a queue, because riders still came and the operator restored supply within hours. At the second, a three-day outage produced an empty forecourt, because riders had already learned not to come. Reliability is not only an operating metric. Once it is lost, demand routes around the site and does not automatically return.

D.3How riders behave in an unreliable system

Across both cases and the wider intercepts, the field team recorded a consistent set of adaptations. They are worth reading as a description of the cost this market imposes on the people least able to absorb it.

• Riders move between stations searching for available batteries.
• Riders call ahead to confirm availability before accepting a job.
• Stations open and close unpredictably, with no published schedule to plan around.
• Riders physically push bikes between locations when a swap fails.
• Income is lost to uncertain swap availability, and customer relationships absorb the delay.

D.4What these observations establish

E

Appendix E

Charging standards and interoperability

Nigeria has no enforced national standard for electric vehicle charging or battery swap. What exists is a set of operator-specific choices. That is the position most markets pass through, and it is the cheapest one to leave behind.

E.1What varies by operator

E.2Charger classes in use

Swap enables sub-ten-minute turnaround, which is what makes high-frequency two-wheeler operation possible. It is the functional equivalent of a fuelling network for that segment, and it is why the two-wheeler ecosystem scaled ahead of the four-wheeler one.

E.3What a standard would unlock

Interoperability is the cheapest intervention available to a regulator here, because it costs nothing to legislate and raises utilisation across every site already built. On the audit evidence, a common access and payment layer alone would address a meaningful part of the discovery friction documented in appendices B and D, without a single new charge point.

We are not in a position to recommend a specific technical standard. That requires an engineering assessment we did not conduct, and it should be led by the industry body with the regulator rather than by any single consultancy. What the field evidence does support is that the absence of any standard is currently costing utilisation at sites that already exist.

F

Appendix F

Model assumptions

Chapter eight is the only modelled chapter in this report. This appendix carries its architecture and its inputs, so any figure in it can be traced, tested or disagreed with.

F.1Model architecture

The model runs 2024 as a baseline through to 2030, in Nigerian naira with US dollar values at the scenario exchange rate, across six vehicle segments. Scenarios are constructed on five independent axes rather than a single dial, which allows combinations a three-point scenario cannot represent.

F.2Scenario drivers, 2030 endpoint values

The full driver set runs to 37 entries; the 22 most material to the figures in chapter eight are shown. The Bear case pairs the highest petrol price with the lowest EV adoption. The two are not contradictory in the model: high pump prices raise the relative attraction of electric operation, but the Bear case also carries the weakest naira, the highest capex inflation and the dearest capital, which suppress the supply of financed assets more than fuel prices lift demand for them.

F.3Vehicle specifications, 2026 baseline

Range is the manufacturer-stated figure for each reference vehicle. Battery capacity multiplied by efficiency will not reproduce it: efficiency is the observed real-world figure used in the energy build, while range is nameplate. Where the two diverge, the model uses efficiency and daily distance, not range. Financed cost is modelled at 70% of acquisition cost throughout. Three different two-wheeler prices appear in the workbook — ₦1.2m in the market sizing build, ₦2.0m as acquisition cost and ₦1.4m as financed cost. None is used directly in a figure quoted in this report.

F.4Battery cost trajectory

US dollars per kWh. Anchored to the BloombergNEF 2025 battery price survey, with a Nigeria premium for import duty, foreign exchange and logistics set on the macro axis. The 2025 survey outturn for BEV packs was $99/kWh, marginally above the 2024 figure. The trajectory here treats that as a pause rather than a reversal; a reader who disagrees should flex the Nigeria premium on the macro axis.

F.5Charger specifications and capital cost

Vehicle-to-chargepoint ratios used as the design basis for the 2030 build: 25 two-wheelers per swap point, 20 three-wheelers, 8 four-wheelers, 4 buses, 6 light commercial, 3 freight. Applied directly these ratios do not reproduce the chargepoint totals in chapter eight, which are the model’s own outputs; we flag the difference rather than smooth it. Note also that the model assumes 85% effective uptime, materially above what the audit in appendix B observed.

F.6Known limitations

Two further limitations belong here.

• No retirement or turnover logic. The parc is a gross running total of everything sold. Vehicles scrapped, written off or exported are not removed, so the 2030 parc is an upper bound on that basis.
• Chargepoint counts are a requirement, not a forecast. They are derived from vehicle-to-chargepoint ratios. They describe what would be needed, not what the market is expected to build.

G

Appendix G

Glossary and abbreviations

G.1Measures as this report uses them

G.2Abbreviations

H

Appendix H

Sources and interview register

H.1Primary research

H.2Interview register

The people and organisations interviewed for this report are listed below. Quotations in the text are attributed at the level each organisation confirmed: some to a named speaker, some to the organisation alone, and one without attribution where the speaker did not grant permission.

H.3Secondary sources

Nigeria macro and banking

• Central Bank of Nigeria, Monetary Policy Decisions, 304th MPC meeting, February 2026 — policy rate path
• Central Bank of Nigeria, Money and Credit Statistics, February 2026 — private sector credit and system liquidity
• National Bureau of Statistics, Foreign Trade Reports 2023 to 2025 — vehicle imports and fleet size estimates
• International Trade Administration, Nigeria Automotive Sector Country Guide — annual vehicle demand and local production
• NNPC and GlobalPetrolPrices — petrol and diesel pump prices
• NERC and distribution company tariff orders — Band A electricity tariff

Nigerian EV ecosystem and policy

• Bridging the Gap: Nigeria E-Mobility Infrastructure to 2060 — long-horizon capacity and capex framing
• Nigeria Energy Transition Plan — generation mix targets used in the energy build
• National Automotive Industry Development Plan (NAIDP), 2023 revision — local assembly targets and import duty structure
• EVN Expo White Paper — battery-to-vehicle ratios and chargepoint density benchmarks
• Electric Vehicle Transition and Green Mobility Bill 2025, and the related policy and legislative review

Global and peer market benchmarks

• BloombergNEF, 2025 Lithium-Ion Battery Price Survey — battery cost trajectory
• International Energy Agency, Global EV Outlook 2024 and 2025 — vehicles per chargepoint benchmarks
• ICCT, Real-World Use Cases for Zero-Emission Trucks — freight efficiency and duty cycles
• RMI, Powering Small-Format Electric Vehicles with Minigrids — two- and three-wheeler technical specifications
• Kenya National Electric Mobility Policy and Kenya Economic Survey 2025 — the peer adoption trajectory anchoring the two-wheeler case
• Thailand BOI EV 3.5 policy, and India FAME II — policy-led adoption comparators
• Africa E-Mobility Report 2025 — continental active EV counts and segment growth rates

H.4How to challenge anything in this report

If a figure here is wrong, we would rather hear it from you than read it in someone else’s correction. Send the figure, the page and what you think it should be to kiko@leke.services. Substantive corrections will be credited in the next edition unless the sender asks otherwise.

—

End matter

Important notice

Scope, limitations and the terms on which this document may be used.