How to sell your software in China: a practical guide for foreign vendors
Entry models, Baidu, WeChat, RMB payments, fapiao, pricing and piracy conversion — what actually has to be true to generate revenue in China, in the order it has to happen.
[NAME] · [ROLE], SystWave · [DATE] · 12 min read
China is one of the largest software markets in the world, and one of the hardest to enter from the outside. Not because buyers don't want your product, but because every layer of the commercial stack is different: the search engine, the social channels, the payment rails, the invoicing rules and the licensing culture. This guide walks through what actually has to be true for a foreign software vendor to generate revenue in China, in the order it has to happen.
Start with a hard question: is there demand, or just a big country?
The most expensive mistake in China entry is assuming that market size equals opportunity. A 1.4-billion-person market is irrelevant if your buyers are 4,000 specialised engineers working in twelve companies.
Before spending anything, answer three questions with data:
- Is anyone searching for your category? Baidu keyword volumes for your product name, your category, and your main competitors will tell you whether there is latent demand or whether you would be creating a category from scratch.
- Who already occupies the space? Local competitors, international competitors with a China presence, and open-source or domestic alternatives. In technical software, a Chinese equivalent often exists at a third of your price.
- How much unlicensed use is there already? Search your product name alongside terms like 破解 (cracked) or 教程 (tutorial) and look at forums, Bilibili and download sites. Heavy pirated usage is not bad news at this stage. It is proof of demand.
Choose your entry model before anything else
Every downstream decision depends on this one. There are three realistic paths.
| Model | What it means | Best when | Main risk |
|---|---|---|---|
| Direct (WFOE) | You set up your own legal entity, hire locally, sell directly | Large budgets, long horizon, China is a strategic pillar | High fixed cost, 6–12 months to operate, HR and compliance burden |
| Reseller network | Local resellers buy licences and sell them on | You already have inbound demand and just need order fulfilment | No demand creation, no control over pricing, positioning or brand |
| Exclusive distribution partner | One local partner handles acquisition, sales, invoicing and support | You want revenue without an entity, and control through one contract | Depends entirely on picking the right partner |
Most mid-sized technical software vendors are better served by the third model for the first two to three years. A WFOE makes sense once China revenue justifies the fixed cost, not before.
Become findable: Baidu, ICP and hosting
Your global website is, functionally, invisible in China. Three things are usually broken at once.
- Google traffic doesn't exist. Baidu dominates. Its ranking logic rewards Chinese-language content, domestic hosting, page speed inside the mainland and site age. Translated pages hosted in Frankfurt will not rank.
- Your site is slow or blocked. Assets loaded from Google Fonts, YouTube embeds, Google Analytics or Google reCAPTCHA will hang or fail behind the Great Firewall. A page that takes eleven seconds to load is a page nobody reads.
- You have no ICP licence. To host a website on a mainland Chinese server, you need an ICP filing (备案), which requires a Chinese legal entity. Without one, you can host in Hong Kong or Singapore as a compromise: faster than Europe, slower than mainland, and weaker for Baidu.
The practical minimum is a Chinese-language site, hosted as close to your users as your legal structure allows, with all blocked third-party scripts stripped out and replaced by local equivalents.
Build presence where your buyers actually are
LinkedIn is not a B2B channel in China. Neither is YouTube, nor X. The channels that matter for technical software are:
- WeChat Official Account. This is your real website. It is where prospects follow you, read your content, contact your team and receive updates. For many Chinese B2B buyers, a company without a WeChat account does not fully exist.
- Bilibili and Douyin. Video tutorials are the dominant learning format for Chinese engineers and students. A tutorial series is often the single highest-ROI acquisition asset for technical software.
- Zhihu and CSDN. Long-form technical Q&A and developer content. Excellent for capturing high-intent research queries and for Baidu visibility.
- Industry WeChat groups. Unglamorous, hard to access from outside, and where a large share of real B2B conversations happen.
The mistake to avoid is translation. Content that performs in China is written by people who understand both the product and the local professional context, not translated from your English blog.
// Intro call
Not sure the demand is there? We'll tell you in 30 minutes.
Make it possible for customers to pay you
This is where a surprising number of otherwise well-executed China entries die.
- Currency and rails. Chinese companies pay in RMB, by domestic bank transfer, Alipay or WeChat Pay. International credit cards are marginal in B2B. A checkout that only accepts Visa and Mastercard converts close to zero.
- Fapiao. The fapiao (发票) is the official tax invoice issued through the state system. Without one, a Chinese company cannot expense the purchase. No fapiao means no purchase order, regardless of how much the end user wants your product. Only a Chinese entity can issue one.
- Procurement rhythm. State-owned enterprises, universities and large private groups run annual budget cycles, often with a spending push near year-end. Design your commercial calendar around theirs, not around your fiscal year.
Getting paid is not an administrative detail in China. It is a feature of your product from the buyer's point of view.
Price for the market, not for a spreadsheet
Straight conversion of your European price list is rarely the right answer, and neither is discounting to the floor.
Consider a China-specific structure: local pricing tiers, education and academic rates, multi-seat packages for design institutes, and bundling with training or support that carries real perceived value. Subscription models face more resistance in markets with a strong perpetual-licence habit, so a hybrid can outperform a pure SaaS move.
The strategic point: your paid competitor in China is often a free cracked copy. You are not competing on price alone. You are competing on legitimacy, support, updates, training and the ability to put a compliant invoice in front of a finance department.
Turn piracy into pipeline
Unlicensed users are the most qualified audience you will ever have. They know your interface, they use your product daily, and they have already chosen it over the alternatives.
A conversion approach that works:
- Make the legal route genuinely easier: local pricing, local payment, instant fapiao, Chinese-language purchase flow
- Attach real value to the legal seat: technical support in Mandarin, verified updates, certified training, official documentation
- Target organisations, not individuals. Companies bidding on tenders or working with international clients need clean licence compliance, and that is a commercial argument, not a legal threat
- Build an academic programme. Students who learn on legal seats become engineers who specify legal seats
Protect your name before you need to
China operates on first-to-file for trademarks. If you have not registered your brand, including a Chinese-character version, someone else can, and you may find yourself unable to use your own name in your largest growth market.
Register early, register in the relevant classes, and choose your Chinese name deliberately rather than letting the market invent one for you. Also map your data obligations: if your software processes user data, cross-border transfer rules under the PIPL may shape your architecture. Take local legal advice on both points; this is one area where general guidance is not enough.
What a realistic first year looks like
- Weeks 1–4: market audit, demand and piracy mapping, competitive and pricing benchmark, go / no-go decision
- Weeks 5–12: entry model and contracts, RMB pricing, ICP filing, Chinese site, WeChat account, payment and fapiao flow, support playbook
- Months 3–6: acquisition engine live (SEO, content, video, webinars), first deals closing, first support tickets handled locally
- Months 6–12: pipeline compounds as content and SEO mature, renewal process established, academic and enterprise programmes started
Revenue in month two is possible. Predictable pipeline takes six months. Anyone promising faster is selling you something.
The five mistakes that cost the most
- Translating instead of localising
- Launching a website with no ICP, no local hosting and blocked scripts
- Offering no way to pay in RMB and no fapiao
- Treating piracy as a legal problem instead of a commercial opportunity
- Choosing a partner who waits for inbound orders instead of building demand
Frequently asked questions
Do I need a Chinese company to sell software in China?
Not necessarily. You can operate through a local distribution partner who holds the entity, issues fapiao, collects RMB and remits your share cross-border. A WFOE becomes worthwhile when China revenue justifies the fixed cost.
How much does entering the Chinese market cost?
It depends on the model. A market audit is a small fixed cost. A distribution partnership is typically revenue-shared, so the fixed cost stays low. A WFOE with local staff is an order of magnitude higher and takes six to twelve months to become operational.
Can I just use my English website with a translation plugin?
It will not rank on Baidu, will likely load slowly or break behind the firewall, and will not convince a Chinese buyer that you are committed to the market.
Is piracy really solvable?
Not eliminable, but very much convertible. A meaningful share of unlicensed organisational users will move to legal seats when the legal route is easy, locally priced, properly invoiced and genuinely supported.
How long before I see revenue?
Setup typically runs 30 to 60 days. First deals usually land in month two or three. Compounding pipeline from search and content builds from month four to six.
[NAME]
[ROLE], SystWave
[AUTHOR BIO — two lines on their background in China software distribution and digital acquisition.]
LinkedIn