Trademark search is the first line of defense for brand safety. A useful search considers similar marks, related goods and services, and whether consumers may confuse their commercial sources.
A question we've been asked hundreds of times
“I've searched this name, and no one has registered it. So it should be okay to use, right?”
After all these years in trademark agency work, this is the question clients ask us most often. Every time we hear it, we understand but also feel concerned — we understand that everyone wants to save time and effort; what concerns us is that the gap between “finding no exact match” and “being safe to register” is enormous.
The overall rejection rate of trademark registration applications in China has long been high, with “prior similarity” being the primary reason for rejection. In other words, many people lose not to identical trademarks, but to trademarks that “look more or less the same.”
01 Why “use first, register later” is the most expensive order
Many companies follow this order for brand development: name the brand → design the logo → launch the product → register only after gaining traction.
This order amplifies risks on three levels:
Level 1: Rejection risk. Using the name before searching means leaving the outcome to chance. Once rejected, all prior brand investment goes to waste.
Level 2: Infringement risk. China follows the “first-to-register” principle. Use itself does not create exclusive rights. If someone else has already registered a similar trademark, the longer you use it and the more you invest, the higher the potential infringement damages.
Level 3: Switching costs. Once packaging, molds, certifications, channel materials, and online store names are finalized, the cost of rebranding rises exponentially over time. It's not just about changing a logo — the entire supply chain has to be adjusted accordingly.
A practical criterion: Any name that will appear on packaging, websites, or contracts should be searched before the first graphic design fee is incurred.
A search costs only a few working days and a small fee, but it brings certainty to these three levels of risk. It's always a worthwhile investment.
02 What is a trademark search really about? Likelihood of confusion
Many people think a trademark search is just “typing in a name and seeing if there's an identical one.” That's not actually the case.
The core of a trademark search is simulating the “likelihood of confusion” judgment made by examiners and courts — would an average consumer, paying ordinary attention, mistake the two trademarks as coming from the same source?
The judgment framework in Chinese practice can be summarized as “two dimensions × three elements × overall observation”:
| Judgment level | Content | Key points |
|---|---|---|
| Dimension 1 | Whether goods/services are similar | Goods in the same class group are presumed similar; related goods across class groups (e.g., clothing and footwear) may also be found similar |
| Dimension 2 | Whether the marks are similar | Comparison of sound, appearance, and meaning; high similarity in any one element poses risk |
| Judgment method | Overall observation + main part comparison | Based on the ordinary attention of average consumers; the dominant distinctive part carries the highest weight |
Sounds a bit abstract? Let's look at examples:
“金典” (Jīndiǎn) and “经典” (Jīngdiǎn) — the characters are different, but the pronunciation is nearly identical. Consumers can't tell them apart over the phone or through word of mouth. This is phonetic similarity.
“康帅傅” (Kāngshuàifù) and “康师傅” (Kāngshīfù) — they look so similar that it's easy to misread them without careful attention. This is visual similarity.
“皇冠” (Huángguān, meaning “crown”) and “CROWN” — one is Chinese and the other is English, but the meaning is exactly the same, pointing to the same concept. This is conceptual similarity.
So a trademark search is not about checking “if anyone uses this exact name,” but about checking “if anyone uses a name that is likely to confuse consumers, while selling similar goods.”
03 Two common pitfalls in self-searching
We've seen too many companies search on the official website themselves, think “it's fine,” only to have their applications rejected and come back to us — it's not that the official website data is inaccurate, but that the self-search approach itself has blind spots.
Pitfall 1: “Adding a prefix/suffix makes it not similar.” Many people think: if someone else uses “XX,” I use “XX Electronics,” “XX Technology,” or “XX International” — that should be different, right? Not necessarily.
Trademark examination focuses on the dominant distinctive part. If the core “XX” is the same, the added “Electronics,” “Technology,” or “International” are all descriptive terms that do not serve a distinguishing function — consumers still remember the “XX.”
A simple example: if your name is “Zhang San” and someone else uses “Zhang San Technology,” people will still think it's a tech company founded by Zhang San, not two different people.
Pitfall 2: “Different classes mean safety.” Another common misconception: if someone registered in Class 25 (clothing), I register in Class 35 (advertising and sales) — that should be fine, right? Also not necessarily.
The Classification of Similar Goods and Services is not set in stone; goods that are related across groups or classes may also be deemed similar. More importantly, well-known trademarks enjoy protection across classes. However, this is not a monopoly across all classes; statutory thresholds apply.
Unregistered well-known trademarks are protected only in respect of identical or similar goods (Article 13(2) of the Trademark Law). Registered well-known trademarks may also prevent an application for goods that are neither identical nor similar if the use is sufficient to mislead the public or harm the interests of the trademark proprietor (Article 13(3)).
In other words, if the opposing trademark is sufficiently well-known, the free-riding is obvious, and the connection between the classes is sufficiently close, an application may still be rejected and legal action may follow, even across different classes. Well-known trademarks are assessed case by case and as required; there is no definitive, once-and-for-all list. Switching classes does not guarantee safety when free-riding on famous brands.
04 Four-step self-search method
Professional searches are not a necessity, but preliminary screening is something everyone can do. The following four steps can be completed using public tools:
Step 1: Identical search. On the CNIPA Trademark Online Service System, search for identical trademarks of the target name in the target class. If an identical prior registration already exists, it's a direct veto — no need to proceed further.
Step 2: Preliminary similarity screening. Decompose the name into variants — pinyin, homophones, common misspellings, English equivalents — and search each one. Focus on the target class and related class groups.
Step 3: Synchronized overseas search. If the brand plans to go overseas, simultaneously conduct preliminary English name searches using the WIPO Global Brand Database and official systems of target countries. Similarity standards overseas are not exactly the same as in China; don't wait until “the Chinese name is available but the English name has to be completely redone.”
Step 4: Professional search and legal opinion. Finding no obstacles in preliminary screening does not mean it can be registered. Similarity judgment, cross-group searches, and investigation of prior unregistered rights (trade names, copyrights, designs) all rely on professional databases and experience. Before formal investment, it is advisable to obtain a written search opinion.
05 How to read a search report: three risk levels, three actions
When you get a search report, don't just look at the words “can it be registered.” A reliable report will grade the risks, with different levels corresponding to different actions:
| Risk level | Typical scenario | Recommended action |
|---|---|---|
| High | Prior identical/highly similar registration with overlapping goods | Change the name, or evaluate cancellation for non-use, invalidation, or assignment of the prior mark before deciding |
| Medium | Similar prior marks exist, but class groups do not fully overlap or the dominant distinctive parts differ | Adjust application strategy: narrow goods, consider combined or separate-element applications, and prepare a rejection review plan |
| Low | No substantial prior obstacles | File as soon as possible (earlier application date wins), and plan core, related, and defensive classes |
A search report is a decision input, not a guarantee. Its value is to compress uncertainty into a manageable range and prepare next steps for each possible outcome.
Closing thoughts
People often ask: “If the search passes, does it guarantee registration?” The honest answer is no.
There are inherent blind spots in searches: prior applications not yet published after filing (the blind period), examiners' individual discretion, oppositions by third parties during the publication period. Even professional searches cannot cover these variables with 100% certainty.
But this is precisely the significance of professional searches — they do not give the illusion of guaranteed approval, but help minimize risks while preparing contingency plans for potential problems.
Trademark rights confirmation is essentially probability management. That is why “search + application strategy + review plan” should be designed as a whole, rather than focusing only on the search step.
After all, the value of a brand never starts from the day it is successfully registered — it starts from the moment you decide to use it and prepare for its risks.
About this article
This article is part of the Trademark Strategy series by Suzhou WISPRO Intellectual Property Agency (Agency Code 32235). The agency provides domestic and international trademark searching, registration, rights confirmation dispute, and brand protection services. Contact us with questions about trademark strategy.