South Africa

The opportunity is an importer/fleet/evidence niche before it is a broad retail aftermarket. Use controlled partner premises, mobile triage and two technical bays; convert to a permanent centre only after stock, demand, platform, legal, waste/fire and economics gates pass.

• 2025 reported 1,088 BEVs and 2,810 PHEVs [naamsa_2025].
• Q1 2026 reported 544 BEVs and 1,277 PHEVs [naamsa_q1_2026].
• The 2025 market totalled 596,818 new vehicles; broad NEVs were 16,716 [market_2025].
• GridCars states it operates more than 350 stations; this is not an official national census [gridcars].

South African “NEV” totals mix BEV, PHEV and non-plug-in HEV. EVS includes only externally chargeable BEV/PHEV in the core pool. The 7,714 denominator adds non-overlapping reported flows for 2024, 2025 and Q1 2026; it is neither active stock nor a forecast and it understates non-reporting brands.

Comparable active-stock history was not located. Reported BEV volume softened in 2025 while PHEVs accelerated, and Q1 2026 plug-ins reached 1,821. The direction is clear; the absolute installed base remains a diligence gate.

PHEVs drove reported growth. BYD’s first reporting month in March 2026 materially changed visibility, with Dolphin Surf, Sealion 5/6 and Shark 6 prominent [byd_march]. This reporting break makes straight year-on-year totals conservative and imperfect.

Cautious/base/high cases apply 25%, 40% and 55% annual growth to the conservative floor over 4.75 years. These are EVS capacity sensitivities, not government or industry forecasts, and must be rebased to active stock.

Operator and industry counts are not directly additive: networks can overlap and stations differ from connectors. Use GridCars’ 350-plus stations as an operator footprint and the approximate 500-public-station figure only as an industry estimate [gridcars] [rubicon]. EVS should diagnose vehicle-side charging and partner for fixed installations.

South Africa’s broad passenger parc averaged 10.5 years in 2022 [parc_age], supporting a mature independent-workshop culture. Plug-ins are much younger, so early demand concentrates in PDI, evidence, charging/12V/HVAC faults, wear and collision isolation rather than mature pack failure.

Do not transfer ICE maintenance economics to EVs. Build utilisation through importer, fleet, insurer and used-EV contracts; add deeper thermal or pack work only after observed fault volume and authorised access.

The model applies 55% annual paid-service incidence and ZAR 14,000 spend to the 7,714 floor, then 42% independent eligibility and 65% single-hub reach. Every rate is an EVS assumption. The result is a conservative service-pool model, not a statistical market forecast.

Included: reported BEV/PHEV flows in 2024, 2025 and Q1 2026. Excluded: HEV/MHEV, pre-2024 plug-ins, unreported brands, attrition, buses, motorcycles, captive warranty work and charger-installation revenue. The true active pool may be higher or lower than the gross floor after attrition and duplication controls.

Importer/new-brand and fleet SLAs come first; insurers/bodyshops and used-EV evidence follow. Retail acquisition begins only after platform coverage and repeatable first-time-fix are proven.

Prioritise BYD’s Dolphin Surf, Atto 3, Sealion 5/6 and Shark 6, followed by Volvo/BMW premium cohorts and demand-led Mercedes, GWM/Haval and Chery/Jaecoo coverage. Toyota’s incoming BEVs are a watchlist, not a tooling commitment [toyota_2026].

OEM networks control warranty, software and parts, while Bosch and RMI workshops offer independent reach. EVS differentiation is audited cross-brand safety, battery evidence, importer/fleet SLAs, collision isolation and disciplined custody—not discount labour.

Johannesburg North–Midrand–Centurion combines corporate fleets, dealer/importer density, premium customers, highways and suitable industrial premises. Cape Town is the second hub; Durban serves the N3 fleet corridor.

No general national cash purchase subsidy was identified. Charging/public-fleet programmes support adoption but must not be treated as consumer rebates [western_cape]. The aftermarket guidelines are strategically positive but do not eliminate OEM access or warranty-fault disputes [competition_guidelines].

Manufacturing incentives, including the reported 150% allowance for qualifying NEV production, are not workshop subsidies [budget_2025]. The pilot case assumes no grant, tax holiday or subsidised finance.

The current company income-tax rate is 27% and standard VAT is 15%; compulsory VAT registration begins at ZAR 2.3 million taxable supplies from 1 April 2026 [cit] [vat] [vat_threshold]. Model values are pre-VAT; payroll, customs, withholding and transfer-pricing advice remains necessary.

Use a small company-controlled technical team with documented competence, medical/rescue readiness and role authorisation. Outsourcing does not remove OHS duties [ohs]. Build an apprenticeship and OEM/importer training pipeline before franchise expansion.

No national workshop-specific EV high-voltage licence was located. That is a risk, not permission: obtain written legal opinions, define authorised roles and secure zoning, occupancy, electrical and fire approvals before intake.

Require identify–isolate–prove dead–lock/tag–verify–rescue procedures, insulated PPE/tools, two-person escalation, controlled keys and evidence logs. No roadside HV work, unsupported energised work or pack opening at launch.

E-waste/EPR sources do not by themselves resolve traction-battery duties [ewaste] [portable_battery_epr]. Obtain a written classification, contract licensed storage/transport/treatment, and approve damaged-battery quarantine and emergency response before accepting a compromised pack.

Refrigerant, heat-pump and pressure-system work requires competent personnel, recovery equipment and current environmental/technical advice. Do not assume foreign qualifications map automatically.

Provide written scope, estimate, parts provenance, limitations and warranty. Preserve OEM warranty boundaries and obtain consent/legal basis for diagnostic and telematics data under the Consumer Protection Act and POPIA [consumer] [popia].

Start company-controlled and asset-light. Franchise only after 12 months of audited safety, first-time-fix, comeback, margin and demand KPIs. Agreements need territory, training, audit, brand, data, parts, insurance and termination controls.

Launch with battery evidence, PDI/campaign support, vehicle-side charging/thermal/12V diagnostics, wear services, mobile fleet triage and collision isolation. Keep fixed chargers with electrical partners and deep pack repair behind a separate gate.

Use two bays, appointment intake, a physically separated quarantine zone, mobile triage and low parts inventory. Publish a platform matrix and price fixed diagnostic/evidence products before broad retail work.

The illustrative base case is 1,200 orders at ZAR 12,500, 56% gross margin and ZAR 7.8 million fixed cost, yielding ZAR 0.6 million EBITDA. The thin base margin reinforces the asset-light recommendation; low case loses ZAR 6.106 million.

The principal risks are stock/reporting uncertainty, warranty captivity, small scale, tool access, battery/fire liability, power resilience, affordability and technician scarcity. Each maps to a measurable pre-lease gate.

The first month buys evidence and legal clarity. The second proves platforms and runs paid controlled trials. The third secures contracts, rebases economics and approves a site only if every gate passes.

No permanent lease until active national plug-in stock reaches 12,000 verified units, at least 6,000 are reachable in Gauteng, 65 monthly jobs are contracted, 70% safe platform coverage is proven, all legal/site gates pass and base EBITDA is positive.

What is the active BEV/PHEV parc by postcode, model and year including non-reporters? Which importer tools and parts can EVS access contractually? Which fire authority will approve quarantine? Which fleets will sign paid SLAs?

Primary policy, tax, safety and regulatory sources were preferred. Registration figures were used through attributed industry reproductions where a directly queryable naamsa file was not indexed. Facts, reported estimates and EVS assumptions are labelled separately; conflicting NEV definitions were reconciled before modelling.

No current official active plug-in stock or national public-charger census was located. The denominator is a conservative gross-registration floor and excludes non-reporting brands. Manufacturer, legal, tax and municipal conditions can change; obtain dated advice immediately before commitment.

The linked registry records publication/access dates, authority and definition notes. Quantitative claims resolve to source IDs; forecasts, sizing, scores, economics and gates resolve to the EVS model.

Conditional go to a company-controlled, asset-light Gauteng validation pilot; defer a full greenfield centre and any franchise rollout. Convert only after the active-stock, contracted-demand, platform, compliance and economics gates pass.