How to Search Using the USPTO’s New Trademark Search Tool
Rachael Dickson | November 13, 2023
Rachael Dickson is a Trademark Attorney at Lloyd & Mousilli and Catalytic Law. Rachael worked as an examining attorney at the USPTO from 2017-2021, where she was awarded a 2020 Customer Service Award. While at the USPTO, Rachael was part of a work group which specialized in reviewing applications to register trademarks with cannabis goods and services. Rachael directed the Intellectual Property and Entrepreneurship Clinic at Suffolk University School of Law as a visiting assistant professor from Fall 2022 – Spring 2024. Her legal article, High Hopes: Cannabis Trademarks at the USPTO, is forthcoming in the Ohio State Technology Law Journal in 2025.
The USPTO is replacing its decades-old search tool TESS (“Trademark Electronic Search System”) completely as of November 30, 2023 with its new search tool. This search tool is currently called “Trademark Search” and accessible online at: https://tmsearch.uspto.gov/search/search-information
Although this new trademark search tool is still in development, it is already a powerful weapon for a trademark attorney to have in their arsenal. It can help identify potential risks that could prevent a trademark from registering, assist with enforcing an existing trademark, and provides a wealth of detailed information for both attorney marketing and academic research.
Why Is This Happening to Us?
TESS originally launched on February 14, 2000 and has long needed an update. However, the USPTO’s rollout of this new tool appears to be motivated more by dire necessity than by a desire to improve the existing search system in any sort of organized or planned way. TESS relies on a backend system and programming language that apparently only one soon-to-retire employee has the ability to maintain. Thus, this tool has been rushed out into primetime without a lot of notice to the trademark attorney community at large; certainly no one seemed to realize that we would lose TESS altogether before the end of 2023 until the USPTO started demonstrating the tool in September.
Advantages and Disadvantages of the New Tool
The new trademark search tool does have several advantages over TESS. It replaces the confusing TESS landing page, which contains three search options with little explanation as to what each can do, with just one page that incorporates all search capabilities into a far more intuitive and navigable format. Search results in the new tool also do not “expire” in the same way that results in TESS did.
This new tool automatically includes the mark drawing, class, and an identification preview for each result. At a glance, you can also identify each mark’s registration and lifecycle status (dead or live, cancelled or abandoned).
TESS’s search language has been replaced with Regular Expressions (also known as RegEx), which is more widely known in the computer science world. Although the field tag format has changed (from [xy], after the search term, to XY:, before the search term), many of the basic field tags remain the same or have an equivalent code that performs the same search functions.
In addition, exports of search results from the new tool offer far more information than TESS’s exports did. These export options are also easier to find: in TESS, you had to know to click a tiny “.csv” link to download the exported results, but in the new tool, an “export” button is clearly visible on the top right of the page.
TESS still does have a few advantages over the new tool, however. Currently the new trademark search tool is far more case sensitive than TESS. If you make an error in the capitalization for a search string, you can get wildly incorrect results; a lowercase field tag can pull up 0 results, while a lowercase operator may bring back millions.
The biggest problem with the new trademark search tool so far is unreliability. Since September, I’ve been running test searches through both TESS and the new tool using search strings from the USPTO’s own training that are supposed to pull up equivalent results. The number of results pulled up by these tools have yet to match exactly, although the difference between the result numbers has decreased over time.
The USPTO has not yet provided an explanation for why the tools are showing different results for equivalent searches. However, they have stated that a lingering legacy system on the backend may show different lifecycle statuses for applications found using the new tool that are different from TSDR (e.g., a mark showing up as dead in TSDR is showing up as live in TESS2). It is possible that this may explain mismatched result numbers as well.
To give the trademark office some credit, they seem to be aware of the limitations of the tool and have been both actively soliciting and responding to feedback. A number of concerns I had with the tool in September have now been remedied.
Searching 101: Dropdown Options
The default search mode for the new tool is “search by all,” which appears to run the typed term through every word field in the record. If you’d like to run a simple search without field tags or sound matching, you can use the dropdown options to search by wordmark, goods and services, owner, serial or registration number, or mark description. You can also use the radio buttons on the left side of the page to filter whether you’d like to look at live marks (registered or pending) and/or dead marks (cancelled or abandoned).
If you use the dropdown options to search the register, you can also specify your search further using a goods/services search box which pops up on the right. Note: This box does not work if you have field tags in your search string, and in fact, the USPTO has now set the box to disappear entirely once field tags are added into the search.
Searching 201: Basic Field Tags
If you’d like to search several fields at once or specify multiple criteria for your search, you will need to use field tags. If you search a string with field tags using any of the drop-down options, the search type will automatically change to “search by field tag.” To work properly, field tags must be typed in all caps, followed by a colon, and then immediately followed by the search term in all lowercase.
Field tags searches can be combined using the Boolean search operators AND, OR, and NOT to include or exclude specific terms or statuses from your search. The full list of field tags is available through the new tool’s Help page under the “Advanced” page and the “Search fields” dropdown.
One of the most basic and commonly used field tags is CM:. This runs a search through the word, translation, and pseudo mark fields. If you’re looking for marks containing the term “unicorn,” you would search CM:unicorn. (This is the equivalent of the TESS search string: unicorn[comb] ).
It’s important to note that once you add field tags into your search, the status filter boxes on the left will no longer work. Instead, you will have to specify the status of the marks you’d like to review using the logical operator AND with additional field tags. You can specify a status in a number of ways.
AND LD:true – live marks.
AND LD:false – dead marks.
AND RN:* – marks that have ever been registered (live or dead)
Searching 201: Phonetic Equivalents and Truncation
In a world full of trademarks such as Krispy Kreme, Froot Loops, Blu-Ray, and Lyft, it’s important to include alternative spellings of a mark in your search. If you ran a search for “crispy crème” alone, you could obviously miss some important conflicting marks which would block your mark from registration. With pattern matching and phonetic equivalents, you can ensure that you’re searching for all possible marks that could sound like yours.
Note: In this new trademark search tool, if you use any sound matching, wildcards, or truncation in your search, you will need to include forward slashes at the beginning and end of your search term.
For example, if you would like to search all live records which have the word CHAT in the wordmark, you’ll want to ensure that you’re searching for all the ways CHAT could be written using alternative spellings.
The sound of the first letter, the soft “c”, could also be represented by S or perhaps Z. You can represent all those possibilities with: [csz]. This tells the system that this one character could be C, S, or Z. The second letter “h” and the fourth letter “t” have no phonetic equivalents, so we can leave those as they are.
The third letter “a”, is a vowel. Generally, it’s best practice to start a search off with allowing for all vowel possibilities at once, as some vowels can sound alike (for example, “i” and “y”), and variations of words sometimes incorporate different vowels as well. You can represent all vowels in a search with: [aeiou].
*Note: While TESS gave you the option to represent some character matches using letters with special meanings, such as “v” for vowels or “c” for consonants, these options are not available in the new trademark search tool. You will need to type in all letters you’d like to search individually.
The USPTO also recommends starting off a search by allowing for the possibility that any 1 character might be replaced by 1-2 characters. You can represent this by using curly brackets immediately after your bracketed characters. You can find a full chart of phonetic equivalents recommended by the USPTO in the new trademark search tool’s help page by clicking the “Advanced” tab, then the “Regular Expressions” drop down option, and the “Phonetic Sounds” drop down option under that.
If you want to ensure that you’ll find any marks that include the term “chat” even if it’s part of another word, or allow for different word forms, you can use truncation before and after the search term. You can accomplish this in the new search tool, by adding .* before and after the term.
So now our search for “chat” looks like this: CM:/.*[csz]{1,2}h[aeiou]{1,2}t.*/ AND LD:true
(this is the equivalent of the TESS search string: *{“csz”1:2}h{v1:2}t*[comb] not dead[ld] )
This search will pull up all live records with word marks that include the term chat, and various variations of it. This search could pull up results such as: chatty, chateau, backchat, but could also pull up less relevant results like chitter, dichotomy, and shuttlecraft.
Although it’s best practice to start searching for the broadest range of variations in the word mark itself, you’re often going to end up with too many search results for you to practically go through. When this occurs, you can assume that the broader search results are going to be less relevant to your case, and you can narrow the search down to something much more similar to your starting mark.
For our “chat” search, we can remove the “s” and “z” options to remove less relevant results like shotgun or shatter. We can also remove all vowels from the search except “a” as this will eliminate less relevant results such as thotchke and chutney. We also can remove the truncation from the front and back of the term to remove other irrelevant results; we can ensure plurals are still searched by adding in [s]{0,1} at the end instead.
Our more narrowed down search looks more like: CM:/chat[s]{0,1}/ AND LD:true
Searching 201: Searching by Class
If you’ve narrowed down your word mark search as much as you can without removing relevant results, you’ll next need to narrow your search results down by classification of goods/services.
Because likelihood of confusion can exist even when the marks at issue have different goods/services in different classes, it’s important not to narrow down your search too fast. Start off with narrowing your search down by coordinated class (CC:), which searches a number of related classes (you can find a chart that details the USPTO’s mapping of related classes here). If our “chat” mark is for clothing, we’ll search CM:/chat[s]{0,1}/ AND LD:true AND CC:025
If you’re still getting too many results to practically go through, you can narrow down to the individual class using IC: as a field tag. If you do narrow down that far, make sure you also include “a b 200” in your class terms to allow for the possibility of collective or certification marks that might have a likelihood of confusion with your mark, and any highly related classes. For example, any goods classes (01-34) should also be searched with 035, the class which contains retail store services. Our narrowed down search for our chat mark would thus, look like CM:/[ckqx]at[s]{0,1}/ AND LD:true AND IC:(025 035 a b 200)
Design Searches
Searching isn’t limited to text and classes; because designs are registrable, the USPTO also has compiled a list of design codes here to enable you to search designs by type. This can help search for any logos or designs that may block the registration of your design.
To search using these codes, add the field tag DC: to any design codes that match your applied-for design. For instance, the design code 05.01.02 is for “Trees or bushes with a generally rounded shape, including deciduous trees.” Although TESS required the removal of the design code’s periods to search properly, the new trademark search tool does not. You can search it either as DC:050102 AND LD:true or DC:05.01.02 AND LD:true
If you’re searching more than one design code at once, it can be very easy to lose track of what number correlates to what design type. In that situation, you can use parentheses to add in notes as to which code is which without affecting the results of the search. DC:(050102 tree) AND DC:(01.01.02 star) AND LD:true
Miscellaneous Useful Searches
If you’re trying to determine if the USPTO considers a specific word descriptive, you can use the field tag DS: to search for marks with disclaimed terms. Remember though: descriptiveness is always determined in context, so you’ll also need to narrow down your search by class, goods, or services to get a reliable answer. While DS:red will bring up any disclaimers of the term “red” in any class, DS:red AND IC:31 will determine whether “red” is considered descriptive of natural agricultural products generally, and DS:red AND GS:(fruit vegetable) will specify the results just to fruit or vegetable goods.
You can search the trademark owner database with the field tags ON: (limited to owner name alone) and OW: (owner with address, for more specificity and granularity) to find all marks held by a competitor. Searching for ON:apple will show all marks owned by any company with Apple in its name, while searching OW:(apple AND cupertino) will show marks owned by Apple (in Cupertino).
Conclusion
This new trademark search tool may seem intimidating, but it can help you learn more about potential obstacles to registration for your mark and about what the landscape of trademarks looks like. Mastering it will provide you with valuable skills to advance your trademark practice.