When a client asks “can we use this?”, get a fast, defensible first read. Usually in under 30 seconds, an examiner-style search of the USPTO register ranks the marks that could block the name, checking sound-alikes, look-alikes, shared words, coordinated classes and foreign equivalents, and records every search it ran. When a name survives the screen, a clearance report adds the written du Pont analysis.
7-day Professional trial, no credit card required.
Tested against the USPTO’s own answer key. In an August 2026 test on 192 applications the USPTO refused under Section 2(d), our search surfaced the examiner’s cited registration in 9 of 10 cases. Method available on request.
Exact match · tmzap
Sounds like · sound key TMSP
Spelled like · letter-pattern similarity
Word parts · across all 48 class codes
Looser spelling · in your classes
Near-misses · one sound off, or run together
Same meaning · by meaning, not spelling
440
different marks found
112
close enough to rank
7.8s
start to finish
Screen a new name against all live USPTO marks, see the closest conflicts ranked by risk and why each one matched, and keep a record of how the search was run.
1:49. Shown on NEXTCEL, for telecommunications services in classes 9 and 38. Transcript · All feature videos
Under 30 sec
Typical time per search
14
Search passes in a typical search
43
Marks rated High or Very High in a typical search, of about 900 ranked
14M+
USPTO filings indexed
Search time: a median of 24.5 seconds across 215 searches from September 10 to October 24, 2026. The weekly median stayed under 30 seconds in each of the 17 weeks from June 29.
Type the name once. The search picks the passes that fit it (exact and near-exact text, pseudo-mark, sound-alike, look-alike, shared words in coordinated classes, foreign equivalents, similar meaning) and runs them in one search. Every result shows which pass found it and why, so you are reading conflicts instead of building queries.
See exactly what was searched, how many marks each pass found and how many made the ranked list. The search is saved to your workspace, so if a client later asks how the name was screened, or a cleared name draws a refusal, the record is already in the file.
Exact match · tmzap
Sounds like · sound key TMSP
Spelled like · letter-pattern similarity
Word parts · across all 48 class codes
Looser spelling · in your classes
Near-misses · one sound off, or run together
Same meaning · by meaning, not spelling
440
different marks found
112
close enough to rank
7.8s
start to finish
Every mark lands in one of four risk tiers, Very High to Low, scored on mark similarity, class overlap and, if you describe them, the goods and services. A typical search ranks about 900 marks and rates 43 of them High or Very High, so you review dozens, not hundreds, before you tell the client whether the name is worth pursuing.
Top of the list
Stormforge, LLC · Class 42 · Registered
NuMe TMS Clinics, PLLC · Class 42 · Registered
Travel Media Group, LLC · Class 42 · Registered
Dongguan Benlei Shangmao Co., Ltd. · Class 9 · Pending
Every search is saved to your recent searches, so you can work through a list of names and come back to any of them later. For a deeper look at any specific match, jump to side-by-side similarity analysis.
Type the proposed mark and pick Nice classes, or let GleanMark suggest them from a description. Describing the goods and services is optional; if you do, each conflict is scored on what it actually covers. List your client's own serial numbers and they are left out of the conflicts.
The search runs the passes that fit the name against the USPTO register, merges duplicates and scores every live mark it finds. Similar dead marks are listed on their own tab as naming history.
Every result shows its risk tier, the passes that found it and a plain-English reason. The page opens with a green, amber or red read on the name. No black box.
Rate the conflicts, export a report for the file, or try the next name on the list. If a name holds up and the client wants a written analysis, start a full clearance report with the name and classes already filled in.
The techniques an examiner uses, run for you, documented and exportable. Once a mark is filed, a watch flags similar new US applications and publications, so you hear about the next conflict early.
Exact and near-exact text, USPTO pseudo-mark readings, sound-alikes, look-alikes, shared words, word order, numerals, foreign equivalents and similar meaning. Exact, sound-alike and look-alike passes run on every name; the rest switch on when the name calls for them. A typical search runs 14 passes, so you do not rebuild a dozen queries by hand.
Finds marks that sound alike when spoken and marks with similar letter patterns, for the whole mark and word by word, using three phonetic codes (Metaphone, Double Metaphone and Soundex) plus letter-pattern similarity. Examiners compare sight and sound; so does this.
Shared words are also searched in commercially related classes, not just the class you filed in. File in Class 25 (clothing) and the search also checks Classes 14 (jewelry), 18 (bags) and 35 (retail services), among others, because examiners do.
Under the Doctrine of Foreign Equivalents, a Spanish or French word is weighed by its English meaning. The search translates foreign words in your mark and also checks English words against their common foreign equivalents.
A How We Searched tab lists every pass that ran, the query it used, how many marks it found and how many made the ranked list, with a plain-English note on the examiner practice behind it (some cite the TMEP section). Every conflict traces back to the passes that caught it. If a client later asks how a name was screened, or a cleared name draws a refusal, the record of what was searched is already in the file.
A one-word compound like GLEANMARK is split into its words (glean + mark) so each is tested on its own. Catches conflicts that a whole-mark search misses.
LUV IT reaches LOVE IT. DOGGIE STYLE reaches OUTBACK HOUND: doggie to dog is a respelling, dog to hound is meaning, the same chain an examiner walks. The search draws on more than 17,000 approved word pairs in about 7,800 equivalence families, mined from pairings examiners relied on in real refusals, from USPTO pseudo-mark readings and from translation statements.
See how dense the field already is around your mark, counted by distinct owners rather than raw marks, because ten registrations from one owner is a family, not a crowded field. The thresholds follow Jack Wolfskin and Juice Generation. Describe the goods and services and each conflict is scored against its actual recitation, not just its class number; more than 99% of live marks have goods text to compare against.
A Top Owners view groups the owners holding several of your conflicts, so you see a portfolio rather than a single hit. Hover any owner for their TTAB record (oppositions, cancellations, live proceedings), with an Aggressive Enforcer flag on owners who file often and usually as plaintiff. Marks the Board has found famous carry a badge with the ruling one click away. Pull two to six marks side by side to compare them.
Rate each conflict yourself, from Very High to Low, or dismiss it. Add notes, filter by risk, match type, class, status, owner or goods overlap, and sort by risk, filing date or name. Your ratings carry into the exported report, so the file reflects your call.
Export a report with the risk assessment, the conflicts grouped by risk with the passes that found each one, crowded-field and goods-overlap findings when they apply, and the search methodology, as PDF, Word or Excel.
The searches you would run yourself on the USPTO’s search system, run for you and written down.
| Task | GM GleanMark Knockout Search | By hand on the USPTO search system |
|---|---|---|
| Search passes | A median of 14 examiner-style passes, chosen for the name and run in one search | You write and run each query yourself |
| Sound-alikes | Automatic: Metaphone, Double Metaphone and Soundex codes | You think up and type the variants |
| Coordinated classes | Added automatically | You look them up and add them to each query |
| Foreign equivalents | Translated automatically, in both directions | You research the translations |
| Ranking | Every live mark scored on mark similarity, class overlap and goods, in four risk tiers | You judge each hit yourself |
| Record of the search | Every pass, its query and its hits, saved with the search | Whatever notes you keep |
| Report | PDF, Word or Excel in one click | You assemble it |
Knockout searches are built for fast screening. When a name needs a document behind your advice, a Clearance Report writes up the same search. Neither covers state registrations or foreign registers such as Madrid/WIPO.
No per-search charges: every plan comes with a monthly allowance, and the 7-day Professional trial includes 25 searches.
2
a month
$0
5
a month
$99/mo
25
per seat a month
$249/seat/mo
Unlimited
$599–$699/seat/mo
When the client is ready to adopt, file or invest, a Clearance Report turns the same search into a document for the file: a written §2(d) du Pont analysis of each strong conflict, a registrability screen, domain, web and social media handle checks, and a plain-English summary for the client. $79 each, with some included in every paid plan.
“GleanMark can streamline the initial retrieval and review process, particularly when screening multiple names in succession. However, the substantive analysis still depends on the complexity of the results.”
The 7-day Professional trial includes 25 knockout searches. No credit card. See what an examiner-style search surfaces before you spend real time on a name.
Start Free Trial