How to actually read Chinese EV reliability data

Chinese brands are arriving in UK showrooms faster than UK reliability data can keep up with them. BYD, MG, Omoda, Leapmotor, XPeng, Zeekr, plus Western-badged but Chinese-built cars like the Polestar 4 and the latest Volvos, are all in the market now, and the established UK and European reliability surveys mostly don't yet have meaningful samples on most of them. By the time they do, you'll be making the buying decision today.

There's a way around the gap that almost nobody writing in English uses: the cars have already been on sale in their home market for years, and China has mature reliability and owner-quality data. The catch is that it's easy to read badly, and a lazy reading will mislead you in specific, predictable ways. Here's how to read it properly, and when not to bother.

What's actually available

Two sources are worth your attention.

J.D. Power China, which runs a New Energy Vehicle Initial Quality Study (NEV-IQS) modelled on its long-standing US studies. It's measured the same way, in problems per 100 vehicles (PP100), and reports both brand-level scores and segment winners. It's the most methodologically robust thing you'll find, and it's broadly comparable across years and brands.

The China Auto Quality Network (车质网, also known by its hotline number 12365), the country's biggest owner-complaint and owner-survey platform. It publishes both structured owner-satisfaction surveys (often "100-owner" snapshot studies) and rolling complaint counts by model. The surveys are useful; the raw complaint counts need handling with care, as we'll come to.

Owner-score platforms like Autohome (汽车之家) and Dongchedi (懂车帝) also exist, with star ratings and free-text owner reviews, and can be useful for sense-checking themes. Their methodologies are looser, so use them to corroborate rather than as the primary source.

The big trap: complaint volumes mislead

Here's the single most useful thing to internalise. BYD regularly tops China's owner-complaint charts. A surface reading of that says BYD is unreliable. Read the actual complaints and a very different picture emerges: the overwhelming driver of BYD's volume is commercial, not mechanical. Owners are angry about steep, repeated price cuts that left recent buyers feeling burned, a phenomenon Chinese owners nickname "backstabbing" (背刺), and they file complaints on the platform as a way of registering their grievance and pressuring for goodwill adjustments. Chinese coverage of these complaint charts says so explicitly: BYD models top the lists because the cars got cheaper, upsetting existing owners, not because of vehicle faults.

That distinction matters enormously, because it's specific to a particular market dynamic that doesn't transfer. The UK doesn't have the same culture of mass complaint filing over pricing, and BYD's UK pricing has been more stable than its mainland pricing. So treating a Chinese complaint-volume ranking as a reliability ranking is like reading Yelp star totals without reading the reviews.

The fix is to filter. Strip out the pricing and dealer-relationship complaints and look at what's left: the genuine quality complaints. On the BYD Seal, for instance, the filtered themes cluster around minor electrical and trim items, software glitches and over-eager driver-assistance behaviour, not the drivetrain or battery. That picture lines up with what both J.D. Power and Western reviewers find. It's a useful, accurate read; the raw chart wasn't.

The transferability caveats

Even good Chinese data is not a UK reliability score. The reasons matter:

• UK cars are export specification. Right-hand drive, often a specific model-year or trim, and frequently a step ahead or behind the version sold at home.
• Software, infotainment and driver-assistance are typically localised for each market, so the experience problems China-market owners complain about may be improved, worse, or simply different in a UK car.
• The price-cut dynamic that inflates BYD's home complaint volumes is largely a China story, less so in the UK.
• Sample sizes, scoring methods and owner expectations differ. A Chinese owner-satisfaction score doesn't convert into a Driver Power score; it's not even on the same scale.

Treat the data as a directional heads-up, not a score. It tells you where to focus your attention on a used or new Chinese-brand car, and what kinds of issues to ask about, rather than telling you how reliable that specific UK example will be.

When it works, and when it doesn't

The Chinese-data approach works best where the brand has high home-market volume, because that's what generates a meaningful body of owner data in the first place. BYD is the obvious example: millions of cars sold at home over years, real J.D. Power coverage, real owner-platform footprints. Use the data here.

It works far less well where the brand is marginal in China. Polestar is a case in point: it builds the Polestar 4 in China, but the company barely sells there (its own retail-network reporting explicitly excludes China), so home-market owner data on the Polestar 4 specifically is negligible. Forcing Chinese-data analysis on a car like that produces a thin, unrepresentative read at best, and a misleading one at worst. The honest move is to say so and lean on other signals: parent-company engineering pedigree, platform siblings that do sell in volume in China, and the brand's track record in markets where it does sell at scale.

The general rule: check whether the brand actually has volume in China before reaching for the home-market data. If it does, the data is genuinely useful. If it doesn't, don't pretend.

A practical framework

If you're cross-shopping a Chinese-brand EV in the UK and want to use the home-market data well, four steps cover it:

1. Start with J.D. Power China. Brand-level NEV-IQS scores and segment results give you a methodologically sound baseline. Note where the brand and the specific model family land relative to the industry average, and which kinds of problems dominate that year (recent studies show issues skewing more toward software and experience than mechanical defects).
2. Read owner-satisfaction surveys, not raw complaint counts. The "100-owner" studies on the China Auto Quality Network filter out the commercial noise and give a usable read on what owners report after living with the car.
3. Filter the complaint platforms. If you do look at complaint counts, separate pricing and dealer disputes from genuine quality issues. That single step transforms the data from misleading to useful.
4. Cross-reference with Western reviews and owner forums. If the China data and the UK or US reviewers agree on something (over-eager driver-assistance, say, or screen-heavy controls), that's a strong signal worth treating as real. If they disagree, weight the source closer to your market.

What you should not do: skip the lot and rely only on UK marketing copy, or, at the other extreme, treat a complaint-chart screenshot as proof a car is unreliable. Both fail buyers in opposite directions.

The bottom line

Chinese-market reliability data is a real and underused signal for UK buyers facing a wave of brands without local track records. Used well, with the price-complaint trap understood, with J.D. Power China weighted over raw complaint volumes, with the transferability caveats kept front of mind, and with discernment about whether a brand actually sells in volume in China, it adds something genuinely useful to the buying decision. Used carelessly, it produces precisely the kind of confident-sounding nonsense that's already filling the internet about Chinese cars. The discipline is the difference.