Devanagari Trademark Search in Nepal: Finding Cross-Script Conflicts Before They Find You
A Devanagari trademark search in Nepal must cross scripts: एभरेष्ट and EVEREST conflict without sharing a letter. The routine, the traps, and the tools.
Nepal's register is bilingual, and its conflicts are too. The mark you are worried about may be filed in Devanagari, romanised three different ways, or hiding behind an English tagline — none of which an exact-match search will surface.
Section 18(1) treats sound-alikes across scripts as deceptively similar. A Devanagari trademark search is therefore not a translation exercise; it is a second and third search in different writing systems.
In short
A Devanagari trademark search in Nepal follows a fixed routine: search the Devanagari form itself, search every plausible romanisation (एभरेष्ट could surface as Everest, Ebarrest or Ebarest), search the meaning-based and phonetic equivalents, then filter by class and read the goods. The DOI's published records complicate this — PDF extraction corrupts text, dates run BS-dominant, and pending unpublished applications stay invisible — which is why structured datasets of 71,000+ records and phonetic matching have become the working method. Filing in Devanagari does not reduce the search burden; registering both scripts is the portfolio answer many Nepali brands take.
Why a Devanagari trademark search breaks exact matching
The law is script-blind: Section 18(1) asks whether marks are identical or deceptively similar for the same or similar goods, and Nepali practice reads sound, appearance and idea across both writing systems. EVEREST and एभरेष्ट share no character and conflict completely.
That makes the search problem two-dimensional. An English-language brand must be searched in Devanagari (how would it be written and pronounced here?), and a Devanagari brand must be searched in romanised English (how would exporters, platforms and young consumers type it?).
The Devanagari search routine, step by step
Search the Devanagari form
The mark as written in Devanagari, including matra and conjunct variations that a filer might have chosen differently.
Search every plausible romanisation
Transliteration standards differ; a Nepali name can surface in the register under several romanised spellings. Search them all.
Search the meaning and the sound
A mark whose meaning translates into an existing English mark — or whose sound matches one across scripts — is a live Section 18(1) candidate. Conceptual similarity is part of Nepalese similarity practice.
Filter by class, read the goods
Class narrows the field under Section 18A; the goods wording decides whether a hit is a conflict or a neighbour.
Record both scripts in the search log
Whatever the search finds, capture the Devanagari and romanised forms, the applicant, the class and the publication date — the log is what an opposition or a clearance opinion will be built from.
Cross-script traps in Nepali searching
The first two rows explain why manual Devanagari searching of raw PDFs is unreliable at scale — and why the structured dataset behind this site's tools (71,000+ published records with phonetic and typo-tolerant matching) has become the practical layer over the Bulletin.
| Trap | Example | The search answer |
|---|---|---|
| Transliteration variance | एभरेष्ट romanised as Everest, Ebarrest, Ebarest | Search multiple romanisations, not one |
| PDF extraction artefacts | Bulletin text corrupted by spacing and glyph errors ('C H E M I C A L S') | Fuzzy and phonetic matching; human review of hits |
| The English name of a Nepali mark | A Devanagari brand whose English tagline conflicts on its own | Search every language the label carries |
| Sound-alike across scripts | A Latin mark that echoes a Devanagari name when spoken | The phonetic layer, always — not just spelling |
| BS-dominant dates | Publication dates read wrongly from the record | Convert once, verify once, store both calendars |
Why the DOI register makes this harder than it should be
The source data is honest but unfriendly: issues of the Industrial Property Bulletin as PDFs, hundreds of records each, dating back to 2006, with dates in Bikram Sambat and text in two scripts. Pending applications that have not yet published are invisible in any case — a search is only as good as the publication record behind it.
The professional workflow absorbs all of this: exact → phonetic → transliteration → class filter → human read. The layers exist precisely because no single pass catches a bilingual register's conflicts.
Filing in Devanagari? The search is still step one
Devanagari filings are ordinary in Nepal and carry practical tips of their own — specimen quality, how the mark displays, how it will be searched by others. But filing in Nepali script does not reduce the English-side risk: an English speaker's later, similar filing is exactly the conflict the routine above is built to catch.
Many Nepali brands resolve the exposure by portfolio: registering both the Devanagari and English forms of the name so the cross-script gap closes from their side. The search routine tells you whether either form is taken before you commit.
Want this handled for your brand?
IP Watch monitors new trademark publications in Nepal and alerts you to potentially conflicting marks — with the context needed to review them.
This article is general information, not legal advice.