22 match types we run on every clearance
Manual trademark search catches the obvious matches. The patterns below are the ones humans miss — but examiners and opposing parties don't. Every Demarka clearance runs all 22 in parallel against 100M+ trademark records.
- Exact match. Identical character-for-character matches across all registries we cover.
- Phonetic letter substitution. EXPRESS → Xpress, Expres, eXpres. Sound-equivalent substitutions examiners treat as identical.
- Cross-script look-alike. COBA → СОВА (Cyrillic). Latin and Cyrillic characters that read identically across scripts.
- Spacing & number-word interchange. 7UP → Seven Up, 7-Up, Seven-Up. Spacing and digit-word swaps count as identical.
- Typographical / transposition. STARBUCKS → Stabrucks, Starbukcs, Starboucks. Typo variants examiners read as conflicts.
- Phonetic compression. GOOGLE → Gugl, Гугл, グーグル. Vowel drops and multi-script transliterations.
- Leetspeak. 2Fast4U → Too Fast For You, 2Fst4U. Number-letter substitutions in modern brand naming.
- N-gram / character rearrangement. nanobanana → banananana, ananabanan. Substring patterns exact-match misses.
- Visual similarity. Letter shapes that look alike at glance: rn / m, cl / d, vv / w.
- Conceptual similarity. Marks with the same meaning in different words — synonyms across markets.
- Translation conflicts. A name that translates into a registered mark in a target-market language.
- Transliteration conflicts. Same name written in a different script (Latin ↔ Cyrillic ↔ Hangul ↔ Kanji).
- Acronym & abbreviation matches. IBM, BMW, KFC — acronyms that read as identical to expanded forms.
- Reverse-order swaps. Word-order swaps in multi-word marks: "Sky Blue" vs "Blue Sky".
- Domain-style compounds. GoFinance, GetWallet — compound forms catching prefix/suffix variants.
- Plural / singular variants. CAT vs CATS, MICRO vs MICROS — number variations.
- Verb-form variants. STREAM, STREAMING, STREAMER — morphological family matches.
- Diacritic / accent variants. CAFE vs CAFÉ — accent-stripping match.
- Common-element decomposition. Marks built from a shared distinctive element with weak qualifiers.
- Disclaimed-word handling. Detecting which element of a multi-word mark is actually distinctive.
- Industry-suffix neutralization. Treating "PAY", "TECH", "WALLET" as weak suffixes in their categories.
- Crowded-field analysis. When many similar marks coexist, identifying the actual distinctive component.