Search a brand name in a trademark database and you will often get hundreds of hits. That number alone tells you very little.
A trademark does not protect a word in the abstract. It protects the word for the goods and services listed in the registration, and those are grouped into classes. Whether a hit matters to you depends first on its class.
This post explains how to read that. It is orientation, not legal advice: whether two uses actually conflict is a question for a trademark attorney or agent.
The Nice Classification in one paragraph
Most trademark offices use the Nice Classification, which has 45 classes: 1 to 34 for goods, 35 to 45 for services. An application names the classes it wants and describes the goods or services in each. One registration can cover several classes.
The classes sellers meet most
| Class | Typical goods |
|---|---|
| 3 | Cosmetics, skincare, cleaning preparations |
| 5 | Dietary supplements, pharmaceutical and hygiene products |
| 8 | Hand tools and cutlery |
| 9 | Electronics, phone accessories, software |
| 11 | Lighting, heating, cooking and cooling appliances |
| 14 | Jewellery and watches |
| 16 | Paper goods and stationery |
| 18 | Bags, luggage and leather goods |
| 20 | Furniture |
| 21 | Kitchen and household utensils |
| 25 | Clothing, footwear and headwear |
| 28 | Toys, games and sporting goods |
| 35 | Retail, advertising and business services |
This table is a rough map. Where a specific product belongs is decided by the office's classification, not by its Amazon category, and some products sit in classes people do not expect. Confirm the class for your product with the office's own classification tool or an adviser.
Seeing a name's class distribution
The trademark statistics endpoint counts the records matching a name, broken down by class:
curl --request POST \
--url https://ecommercedataapi.com/v1/global/brand/stats \
--header "Content-Type: application/json" \
--header "X-API-Key: ${ECOMMERCE_DATA_API_KEY}" \
--data '{
"office": ["US", "EM", "JP"],
"text": "AURORA",
"returnFields": "office,status,niceClass"
}'On 12 October 2026 this returned 2,679 records across the three offices. The ten largest classes:
| Class | Records | Class | Records |
|---|---|---|---|
| 9 | 525 | 41 | 173 |
| 42 | 243 | 3 | 137 |
| 5 | 236 | 34 | 137 |
| 35 | 225 | 16 | 130 |
| 11 | 202 | 10 | 129 |
"2,679 hits" and "202 records in class 11" are different facts. A seller of lamps cares about the second. The statuses matter too: in the same result, 1,419 records were Ended and 136 Expired, so many of those records are no longer in force.
Because one registration can cover several classes, class counts do not add up to the record total.
Trademark statistics endpointCounts of matching trademark records by office, status, class, applicant and yearNarrowing a search to your class
Once you know the class, filter the search to it and to the statuses that are still alive:
curl --request POST \
--url https://ecommercedataapi.com/v1/global/brand/list \
--header "Content-Type: application/json" \
--header "X-API-Key: ${ECOMMERCE_DATA_API_KEY}" \
--data '{
"office": ["US", "EM", "JP"],
"text": "AURORA",
"status": ["Registered", "Pending"],
"niceClass": [11],
"page": 1,
"size": 20
}'Each record lists its niceClass array, applicant, dates and status. The parameters and fields are covered in the trademark search API reference, and the full screening sequence in checking trademarks before you list.
Three mistakes
1. Checking only your own class. Goods in different classes can still be treated as related, and well-known marks can be protected beyond their classes. Look at neighbouring classes, then ask an adviser about the ones that worry you.
2. Ignoring class 35. Retail and online-store services sit there. Brands that sell through their own storefront often register in 35 as well as their product class.
3. Reading a different class as a green light. A different class lowers the risk. It does not remove it. The API reports classes; it does not decide conflict.
What this cannot tell you
Whether your use would infringe. That depends on similarity of the marks and of the goods, judged under the law of each jurisdiction.
Whether the register is complete. Data lags the offices, and coverage varies by office.
Whether a marketplace brand registry will accept you. That has its own requirements.
Questions
How many trademark classes are there? 45 under the Nice Classification: 34 for goods and 11 for services.
Can two companies own the same name? Yes, when they are registered for different goods or services and are not likely to be confused. The search result for one common word above shows hundreds of live records.
Which class do I need for my product? The table above is a starting point. Confirm it with the trademark office's classification tool or a professional before you file.
Can I filter a trademark search by class?
Yes. Pass class numbers in niceClass.