Framed around trademark examination practice, this guide systematically explains the complete logic by which examiners judge trademark similarity: the judging subject, comparison methods, similarity rules for each type of mark, the linkage with similarity of goods, and how to use this logic before filing to anticipate risk.
1. The Subject Standard: The Ordinary Attention of the Relevant Public
Similarity judgment is not a fine expert appraisal; it simulates the cognitive state of the relevant public when selecting goods. The relevant public means the consumers and operators connected with the goods or services. Ordinary attention means that observation is general rather than meticulous, and judgment relies on remembered impressions rather than side-by-side comparison of the actual marks.
Two important corollaries follow. First, subtle differences do not change a similarity conclusion: adding or removing a single stroke or changing an individual radical leaves the public unable to distinguish the marks from memory, and similarity is still found. Second, how a mark is called and remembered carries great weight: a catchy homophonic mark with the same pronunciation can easily be lumped together with the earlier mark by the public even if the characters differ.
2. Basic Comparison Methods
Overall comparison: first look at whether the overall impression the two marks leave on the public is close; where the overall visual impressions clearly differ, similarity may not be found.
Dominant-portion comparison: then focus on the distinctive identifying portion of the mark. In word marks, the beginning of a word, a coined word, or a word with concrete meaning is usually the dominant portion; meaningless function words and generic suffixes (such as "之家", "优品", "官方") are weakly distinctive and basically do not affect the similarity conclusion. In combined marks, the word portion is generally the dominant portion for public calling and memory.
Isolated comparison: the judgment is made with the marks observed separately, asking whether the public would be confused relying on impressions—not by placing the two marks side by side to spot differences.
3. Similarity Rules for Word Marks
Similarity in character form: look-alike characters (such as "太白" versus "大白"), added or removed strokes, or a changed structure with an overall similar appearance lead to a similarity finding.
Similarity in pronunciation: same sound in different characters (such as "娃哈哈" versus "娃娃哈") leads to a similarity finding in principle; where the pronunciation is only partially close and the meaning and character form differ greatly, similarity may not be found.
Similarity in meaning: synonyms, near-synonyms, and Chinese-foreign pairs with stable corresponding meanings ("太阳" and "SUN") that readily trigger public association lead to a similarity finding—except where the foreign word's meaning is obscure and public recognition is low.
Word order and added characters: where reversing the character order still leaves the appearance and pronunciation close, similarity is found; where the mark fully incorporates a prior trademark and merely adds a weakly distinctive modifier (such as "新", "大", "之", "牌"), similarity is found; where the added portion creates a new, definite meaning and differentiates the whole, similarity may not be found.
Meaningless letter combinations: judged by the arrangement of letters, overall appearance, and how they are called; similar appearance means a similarity finding.
4. Similarity Rules for Device Marks
Device comparison looks at compositional elements, design style, and overall visual effect. The three carry different weights: identical or highly similar compositional elements (for example, two realistic depictions of a running horse) are dangerous even if the styles differ; where compositional elements differ but the design style and overall impression are identical, the case is judged individually. Where the public can read a definite meaning directly from a device, that meaning also enters the comparison—for example, an "apple device" and the word "苹果" (apple) may be found similar because their meanings correspond.
5. Similarity Rules for Combined Marks
Examination of a word-device combined mark is "element-by-element examination": if either the word element or the device element is similar to a prior trademark, the entire application is rejected. A combined application therefore carries higher risk than separate filings.
Similarity of the word portion is the most common ground of rejection in practice, because words are the public's primary way of calling a mark. Conversely, a combination whose words differ but whose device is highly similar will likewise be blocked. For a Chinese-foreign combined mark, similarity of the Chinese portion to a prior Chinese mark, or of the foreign portion to a prior foreign mark, both constitute obstacles.
6. The Linkage Between Similarity of Marks and Similarity of Goods
Likelihood of confusion is determined jointly by the degree of similarity of the marks and the degree of similarity of the goods, and the two rise and fall together. Identical trademarks on dissimilar goods usually do not conflict; highly similar trademarks on goods of ordinary relatedness may still be rejected. The more renowned the prior trademark, the more the linkage scale tips toward similarity: for a well-known mark, confusion may be found even across a wider span of goods.
Similarity of goods is judged on the basis of the Classification of Similar Goods and Services: goods in the same group are similar in principle, cross-group and cross-class relationships noted in the annotations take precedence, and in individual cases function, use, channels, and consumers are considered comprehensively.
7. How to Use This Logic for a Pre-Filing Self-Check
Step one, put yourself in the examiner's shoes: using isolated comparison, judge by first impression whether your mark and each suspected prior trademark would be confused. Step two, find the dominant portion: identify the distinctive identifying portion of your mark and confirm whether the dominant portion of the prior trademark overlaps with it. Step three, sieve across three dimensions: check character form, pronunciation, and meaning each once; high similarity on any dimension lands the mark on the high-risk list. Step four, run the linkage check: grade the degree of similarity together with the similarity of the goods, and abandon outright where high similarity meets similar goods. Step five, for a combined mark, run an element-split check, passing each element through the prior-rights database individually.
8. The Response Interface After a Rejection
After a similarity-based rejection, the room for argument in a review of rejection lies in: differences in overall appearance and meaning, non-overlap of dominant portions, dissimilarity of goods, changes in the legal status of the cited mark (non-use cancellation, invalidation, or expiry without renewal), and market differentiation established through use of the two marks. The success rate of a review is strongly correlated with the degree of similarity; the more rigorous the pre-filing self-grading, the less passive the after-the-fact remedy.
9. How to Get It Handled
If you need assistance with the above procedures, you can book a trademark registration assessment with a record-filed agent on MyTMBee, with full assistance from document preparation to filing follow-up.