Researching China: assess a segment without relying on one number
Research a defined segment in China through data definitions, channels, conversations and support requirements. Build a decision that does not depend on one large figure.
Article language: English

In this article
A whole country on one slide can look like a ready-made plan until someone has to choose whom to approach first. Base research into China's market on a defined customer group, a use for the product and a route to that group. A large national figure can provide context, but it will not tell you who will buy your solution or what needs checking before an initial trial. Combine data of the right scope with an examination of offers, conversations and an assessment of your own ability to support customers. The result should be a choice of next action with its conditions made explicit, rather than a single number standing in for the decision.
This does not mean abandoning data. Each piece of information needs an appropriate role. Statistics describe a defined phenomenon, a catalogue shows a visible offer, and a conversation helps explain a particular case. The process below shows how to combine them without confusing national scale with demand you can actually reach. The example applications and decisions are hypothetical; the current factual claims point to specific source materials.
Define the decision before looking for market size
Establish what the company needs to choose: a product application, a type of partner, an area for initial conversations or a limited trial. Each decision calls for different data. If you are considering a component for a particular process, total retail sales do not describe its potential buyers. With liquid cooling from China (in Polish), establish whether you need a server component or part of a facility's infrastructure. A shared sector label does not settle the scope. If you need a channel for a finished product, the number of manufacturing plants may not answer your question about access to customers.
Use the brief for research into Asian markets (in Polish) to connect the question to the result you need. That might mean mapping an application, setting criteria for selecting organisations and planning how to check the most important unknown. Do not commission a general overview of China and expect a complete sales model to appear in its final paragraph. That shortcut saves space in the enquiry but usually leaves the team with the unresolved problem of choosing a scope.
Separate national context from the segment
Describe the segment through the customer's task, conditions of use and purchasing process. Then establish which public information might help identify it. Industry, company size and location are useful when they change the question or the way you would serve the customer. They do not have to produce dozens of subgroups at once. Initially, you need a definition that another person can apply to a company profile. The Chinese photovoltaic value chain (in Polish) illustrates this well: material, cell, module and installation point to different places to look for buyers and different units in the data.
Segmenting B2B markets in Asia (in Polish) helps distinguish a statistical category from a group sharing a similar need. Check a few cases before collecting data widely. If companies qualify only because they use the same word in their descriptions, return to the application. A large group with unclear criteria is no more useful than a small one. It is simply harder to review.
Read the definition alongside the indicator
The NBS release on retail sales for January-September 2026 describes, among other changes, expanded statistical coverage of online services and changes to indicator names. The agency states that the new indicator for online retail sales of goods and services is not comparable with the former online retail sales indicator. This is a specific example of why a name and a methodological note can change the meaning of a comparison.
Before using a series, record its unit, product scope, period, geographical area and method. Check whether the comparison concerns the same phenomenon and whether the source identifies a change in coverage. Do not apply a broad result to your own segment without justification. A figure can be correct yet unsuitable for your decision. In source-based research, that is information about the material's limits, rather than a contradiction.
Check who the survey covers
The introduction to the domestic trade chapter of the NBS 2024 yearbook distinguishes populations above a designated size from those below it, as well as comprehensive reporting from sample surveys. It is an example of the method described for a particular item in the 2024 yearbook. It does not automatically establish the rules for every table published in 2026. Do not treat a table covering a selected population as a straightforward list of all companies. Check the definitions and current coverage of each new dataset.
In your spreadsheet, distinguish the number of organisations from the number of sites, sales outlets or other units. If sources use different objects, do not add them together automatically. For your own list, apply the approach described in organising company databases from Asia (in Polish). A clear unit is necessary before you can discuss how complete the collected set is. Without it, even careful counting can produce a result nobody can explain.
Choose the area for the initial investigation
Rather than comparing every possible location immediately, define a criterion for choosing an area. It might be its connection to the application, access to people you can interview or the practical ability to support a trial. Justify the choice using sources and the project's needs. Do not present convenience for the researcher as market attractiveness. An easily accessible browser tab does not make an easily accessible market. Available material does not establish that the right customer group is located there.
Also record what you will exclude. A narrower study is useful when its boundary remains visible in the recommendation. Similar clarity helps compare options when choosing a first market in Asia (in Polish). If one proposal concerns an entire country and another a single application in a particular channel, they are not equivalent alternatives. Agree on the question before evaluating the answers.
Identify the channel and each organisation's role
A company you find may be a manufacturer, reseller, integrator or customer. Check its role before giving it a place in the market-entry plan. An offer visible on a platform shows one way of presenting a product; it does not automatically describe the entire channel. For each item, note what it establishes and what remains an assumption. Similar names cannot stand in for a missing connection between organisations.
Connect a comparison of distribution and direct sales (in Polish) to the tasks involved: who will explain the product, run a trial, answer a question and handle a support request? The channel is part of how value reaches the customer. Finding an intermediary does not resolve those issues. You need to check the scope of its role and which responsibilities remain with your team.

Bring together three types of evidence
For the selected application, prepare three separate sets: existing data, public descriptions of solutions, and information from conversations or trials. Record how each was obtained and its limitations. Then check whether they concern the same segment. A general statistic, one company profile and an interviewee's opinion may all be useful, but they do not necessarily support a single statement about the scale of demand.
Testing demand in an Asian market (in Polish) means moving from an assumption to checking a specified action. A request for specifications, willingness to take part in a trial and a completed purchase are different events. Do not combine them into one category of interest without explaining the distinction. A smaller number of clearly described signals will be more useful in the report than a long list in which every row means something different.
Practise comparing two applications
Imagine a supplier of a packaging system. In this exercise, it is considering factories replacing existing equipment and integrators building new workstations. These are two research tasks. The first group may require an understanding of the current process and the conditions for replacement. For the second, you need to check the component's role in the project and the information required for integration. No general retail sales indicator can, by itself, settle which group fits the product.
For both applications, record the evidence needed and a possible trial. Then compare access to interviewees and the cost of preparing materials. If one route requires data the company does not yet have, include that dependency. Do not call the other route a better market solely because it is easier to investigate. You can, however, justify the order of research: first a trial the current team can carry out, then a wider assessment of the alternative.
After the conversations, return to the group definitions. Perhaps some factories buy only complete solutions, while some integrators do not select components themselves. These are hypothetical answers illustrating how a criterion can change. Revise it and keep the reason for the revision. Do not preserve an old segmentation merely to keep the original spreadsheet looking tidy. The spreadsheet should serve the research, not the other way around.
Check whether your own offer is feasible
Alongside information about buyers, list your own requirements: the language of materials, availability of a technical specialist, arrangements for a trial and handling questions. A missing element can change the order of actions. You do not need to build a complete local organisation immediately. First establish what a limited investigation actually requires and what can be confirmed before making larger commitments.
Preparing an Asian company's entry into Poland (in Polish) shows the same principle from the other direction: materials, channel and support must form a workable process. When researching China, assess your own readiness as honestly as the market data. The scale of the surroundings will not fill in a missing specification or answer a technical enquiry. Those tasks still need an owner.
Test the decision under two different estimates of scale
Prepare two working scenarios without giving them false precision. In the first, the reachable group of buyers turns out to be smaller than expected. In the second, more organisations meet the definition. Check whether you can run the same limited trial in both cases. Today's most important unknown may be how a product is evaluated, rather than the total number of potential buyers. The next stage should then investigate that process.
If the choice changes between scenarios, identify the point on which it depends. Make clear whether the issue is the cost of reaching customers, support availability or another part of the plan. Saying that a larger market looks better does not explain the decision. The team should be able to show which assumption needs checking before expanding its activity. A further report or conversation will then have a defined task, rather than becoming another attachment to an unchanged plan.
Keep the option of postponing the project as well. That can be justified if you cannot currently assess the application, prepare the materials or reach someone in the right role. Do not treat it as a judgement on China as a whole. It is a decision about a particular task, given the company's current knowledge and resources. Add a condition for returning to it: completing documentation, being able to run a test or clarifying an important unknown. That gives you a reason to revisit the subject.
Prepare questions for the analyst
Ask the author to identify which data describe the wider environment and which concern the selected application. Then choose one recommendation and trace its basis. Does it come from a statistic, a company profile, a conversation or an assumption made by the team? Does the material cover the right unit and period? What answer was obtained, and what remains a hypothesis? Following that chain tests the reasoning more effectively than simply asking how many sources were used.
Finally, ask about an alternative explanation. No reply may reflect an unclear message rather than no need for the product. An offer you find may concern a different configuration. A difference in values may come from a change in statistical coverage. The author does not have to resolve every possibility, but should show which were checked and why a particular next step was chosen. That provides a basis for discussing the decision rather than arguing over whether the market is simply large.
Finish with a recommendation that can withstand a change in one number
Prepare a preferred option and the conditions that would change the choice. State what remains justified under different reasonable estimates of the segment's scale. If the decision depends entirely on a single uncertain value, identify that dependency and propose a way to investigate it. Do not create an artificial range merely to make the report look cautious. Every value and assumption needs its own basis.
When closing research into an Asian market (in Polish), connect the conclusion to the next action, its owner and the evaluation criterion. For Aeneia Oriens, a useful starting point is the product, a working customer group and one decision that requires investigation. Research should provide an explainable basis for action, without hiding behind the size of the whole country.







