Stage one: prove someone is searching
Before spending on anything else, confirm that the problem your product solves is discussed in Chinese, by people who can buy software. We check this against real search behaviour and platform activity rather than against a market report.
- Which Chinese terms describe your category, and which of them have commercial intent.
- Whether the questions your buyers ask are already being answered — and by whom.
- Whether competitor or adjacent products are visible in Chinese search results.
If the answer is that nobody is looking, no amount of content will fix it. That is worth knowing in week one rather than month nine.
Stage two: build a Chinese surface
Demand needs somewhere to land. We build a minimal Chinese surface: a product explanation written for Chinese practitioners, a presence on the one or two channels where your buyers actually spend time, and a way to capture interest that does not require a local phone number or a local bank account.
Which channels matter depends on the product. Infrastructure and developer tools concentrate on Zhihu and technical communities. Design and collaboration tools tend to work on Xiaohongshu. Industrial software is closer to WeChat and industry events. We say plainly which one we think applies to you.
Stage three: decide on structure
Only after interest is measurable does the structural question become worth answering: local entity or not, distributor or direct, RMB invoicing or USD. These are legal and financial decisions that depend on your jurisdiction and your risk tolerance.
We are not a law firm and we do not give legal advice. What we do is give you the operating evidence — what channels produce, what buyers ask for, what a local partner would need to handle — so that the conversation with your counsel and your finance team starts from facts.
Where foreign vendors usually get stuck
- Translated positioning. The English positioning rarely survives translation intact; the buying criteria in China are often different.
- Wrong first channel. Launching on a consumer platform for a product sold to engineers wastes the first quarter.
- No owner. Content published by whoever has time stops within two months.
- Access friction. Web properties hosted outside China load slowly or intermittently, which quietly kills conversion.
- Compliance guesswork. Assuming rules are either stricter or looser than they are, in both directions.