Information date: 2 September 2026 — China’s Ministry of Industry and Information Technology reported 1.561 million new-energy vehicle sales in July 2026, up 23.7% year on year and equal to 60.4% of all vehicle sales that month. A foreign component, charging, software or after-sales provider can treat that as evidence of a large active sector. It cannot treat the national share as proof that its offer will win customers in a selected city.
Start with the verified fact
The same MIIT release records total vehicle sales of 2.584 million in July, slightly below the previous year. NEV production was 1.576 million. For January through July, NEV sales reached 9.007 million, up 9.6%, and represented 51.2% of total vehicle sales. July vehicle exports exceeded one million and NEV exports were 553,000, up 150% year on year.
These are producer and market totals, not a forecast for one model, technology or service. The July share can be affected by production, incentives, product cycles and the denominator. It does not state which cities, vehicle segments or purchasers would adopt a foreign entrant’s product, and it does not show the entrant’s accessible revenue.
How the mechanism reaches an operating decision
An EV offer reaches revenue through vehicle-platform qualification, technical integration, product compliance, procurement approval, local service and payment. A component maker may face long validation before mass production. A software provider may need vehicle data access and cybersecurity review. A charging service may depend on property, grid and utilisation economics. One national sales statistic does not remove any of those gates.
The useful case is therefore a narrow pilot. It identifies one customer problem, a specific vehicle or fleet, a city, the integration owner, the success metric and the commercial decision after testing. The pilot must be designed to reveal a failure cheaply; a demonstration with no procurement owner or acceptance rule only creates publicity.
Choose whether to proceed, redesign or pause
Proceed to a paid or tightly scoped pilot when the customer confirms the use case, provides access needed for testing, and identifies technical and purchasing decision makers. Redesign when the problem is real but the original product needs localisation, different data architecture or a different vehicle segment. Pause when access, licence, safety evidence or ownership of results cannot be resolved before substantial spending.
Do not choose the pilot city only from registration volume. Compare the customer cluster, regulatory pathway, partner capability, service distance and the ability to measure a commercial outcome. A smaller fleet with a committed operator can produce stronger evidence than a large market with no accountable buyer.
Turn the decision into an evidence file
- Define the offer at SKU or software-module level and state the exact operating problem it changes.
- Select one buyer, one city, one vehicle or fleet context, and one measurable baseline such as downtime, energy use or service response.
- Map mandatory standards, testing, cybersecurity, data, import and licensing dependencies to named owners before equipment is shipped.
- Agree pilot access, duration, sample size, failure handling, intellectual property, data retention and a decision date in writing.
- Calculate pilot contribution after localisation, engineering, travel, hardware, tax and collection costs rather than reporting gross contract value.
- Scale only if the acceptance evidence is repeatable; retain the national MIIT figures as context, not as the pilot result.
Run one proportionate review
For China EV launch case, review the items that could change the commercial conclusion: the current primary source, the applicable entity and date, the owner of each open task, and the evidence required before money or customer commitments become irreversible. A wording preference is not a control failure. Correct a draft only when a factual error, unsupported claim, missing source, unsuitable taxonomy or unusable action would mislead the reader.
For China EV launch case, keep a compact record of the source URL, access date, assumptions, responsible person and next review trigger. When a rule, product or market figure changes, update the affected row and decision instead of rebuilding evidence that remains valid. This preserves traceability while keeping the process economical.
Boundary of the conclusion
This case framework does not predict Chinese EV demand or recommend an investment. MIIT’s figures are national aggregates and may be revised or interpreted with other datasets. Vehicle sales, registrations, fleet deployment and component orders are not interchangeable measures.
Safety-critical components, autonomous functions, batteries, mapping and personal or important data require specialist review. A pilot approval is not a national operating licence, and a successful technical test is not proof of profitable scale.
