Hi readers,
China is, depending on who you ask, either one of the most exciting growth opportunities in companion animal health, or one of the most frustrating market to model. Often, it’s both.
China’s pet population - especially cats - is large and growing, particularly in urban centres. Veterinary infrastructure is expanding. Premium product adoption is moving faster than most Western markets did at similar stages of development. Most global animal health companies already have a substantial presence in the country.
However, during conversations with executives at this year’s AHNTI conference in London, there was also palpable frustration over a major area: the data.
Not because good data doesn't exist, but because the Chinese market generates several distinct categories of data with very different levels of reliability.
However, this doesn’t make the China feline market impossible to model. It does mean that a slightly different approach needs to be taken to the data. This week’s Feline Business Brief sets out to provide a practical guide to interpreting China data with confidence.
What we’re watching
Why Chinese market data behaves differently
The starting point is data distribution.
A substantial proportion of veterinary product sales in China move through channels that don't generate clean, auditable transaction records:
Regional distributors with variable reporting practices;
Rapidly growing e-commerce platforms; and
Informal supply networks that sit outside conventional trade tracking.
Sales figures based on these data are therefore often reconstructions rather than primary records, put together through successive rounds of estimation. Even companies with strong local presence are frequently working from triangulated approximations, rather than verified data.
Regulatory filing data is more reliable, but is often released after a significant time lag. This means the picture the data describe reflects commercial reality from twelve to eighteen months earlier. Given how fast the Chinese companion animal health market moves, this renders the data less useful when modelling the commercial environment. For example, a product category that looks nascent in regulatory data may already be a competitive commercial space in reality.
Survey data introduces a different problem. Pet owner and veterinary surveys in China systematically oversample urban, digitally engaged, premium-oriented respondents - the population most likely to participate in formal research.
This group is not representative of the broader Chinese market, which remains heterogeneous across geographies, income levels, and veterinary infrastructure. Treatment compliance rates, premium product adoption, and veterinary visit frequency derived from survey data will almost always skew high as a result.
Finally, KOL-derived intelligence also requires particular care. Chinese veterinary opinion leaders frequently carry commercial relationships that are not formally disclosed, which affects how their clinical usage estimates and market assessments should be weighted in any synthesis.
A tiered approach to source confidence
Rather than treating all incoming data with uniform scepticism or uniform acceptance, it is more useful to adopt a tiered source hierarchy.
At the more reliable end:
Customs and import/export trade data: Independently generated, consistently structured, and not subject to the relatively opacity of domestic sales reporting. It captures volume rather than value, which limits its utility, but can be used to cross-check other data sources.
Regulatory approval and filing data: Reliable, despite the time lag problem (which can be adjusted for). It is particularly useful for tracking competitive pipeline activity and category development.
More indicative than quantitative:
Distributor and supply chain interviews: This is more directional and qualitative. Useful for understanding market dynamics and channel behaviour, but not appropriate as a basis for quantitative estimates without cross-checks.
Survey data (pet owner, veterinary, and KOL-derived): This should be treated as indicative rather than quantitative. It is most useful for identifying directional trends and hypothesis generation, least useful when precise adoption or utilisation figures are required.
What this means for market models
Practical implications:
Chinese market estimates should carry explicit confidence intervals, not point estimates.
Internal strategy discussions should be clear about which source tier is driving which assumption.
Even if the headline numbers look similar, a total addressable market (TAM) model constructed mainly on distributor interview data and pet owner surveys will be substantially different from a model based on trade data and regulatory filings.
In short: The limiting factor on building reliable TAM models is usually not data volume but interpretive framework: a structured approach to data source weighting.
This would allow global animal health companies to engage critically with the data produced by local partners in China, rather than accepting or discarding the data wholesale.
Feline Business Brief provides competitive intelligence on the global cat sector. We analyse early signals, emerging risks and structural shifts across feline health, therapeutics, diagnostics and technology. Learn more here.
We provide:
Global feline pipeline tracking
Market sizing
Competitive landscape mapping
Bespoke research
