How to Localize a Product Title for Coupang Search (Beyond Translation)

Why a translated title almost never ranks

A translated title answers the question "what is this product called in Korean?" A localized title answers a different question: "what words does a Korean shopper actually type into Coupang's search bar to find this?" Those two answers overlap less than you'd expect.

Translation engines pick the dictionary-correct term. Korean shoppers frequently search with a loanword, an abbreviation, a slang compound, or a spec number instead. If your title contains only the dictionary term, you are invisible for the query that carries the volume — and on Coupang, where the buy box and search ranking are tightly coupled, invisibility in search is close to invisibility overall.

The fix is not more Korean fluency. It is treating the title as a keyword container with a fixed budget of characters, then deciding which demand you want to capture with each slot.

• Translation optimizes for correctness; localization optimizes for query match.
• A single wrong term choice (e.g. native Korean where shoppers use a loanword) can cost you an entire keyword cluster.
• Coupang's search is unforgiving about words that simply aren't in your title — assume no semantic rescue.

Step 1: Find the word shoppers actually type

Before writing anything, build a short list of candidate terms for your product and check which ones carry real search demand. Korean product vocabulary usually has two or three competing forms: a Sino-Korean or native word, an English loanword written in Hangul, and sometimes a shortened compound that only exists in shopping contexts.

Take a yoga mat. 요가매트 (yoga mat, loanword) is what people search. A more "correct" descriptive phrase would get almost nothing. Or a humidifier: 가습기 (humidifier, Sino-Korean) dominates, and the loanword form is dead. There is no rule — you have to check each one.

This is the step most foreign sellers skip because checking Korean search volume in Korean-language tools is painful. SeoulRank exists partly for this: it surfaces Korean keyword volume and variants with English glosses so you can compare candidates without guessing.

• List 5–10 candidate head terms per product, including loanword and native forms.
• Check demand for each, then check what the top-ranking listings for that term actually use in their titles.
• Note modifier keywords shoppers attach: 무선 (wireless), 대용량 (large capacity), 휴대용 (portable), 인테리어 (decor/aesthetic).
• Watch for shopping-specific compounds that no dictionary lists — e.g. 캠핑용 (for camping) as a productized modifier.
• Record singular spellings: Korean has no plural marker in search, so don't try to cover both.

Step 2: Order matters — front-load the head keyword cluster

Coupang titles get truncated in mobile listings, and the earliest words carry the most weight for both the algorithm and the human scanning a grid of thumbnails. Put the brand and the core category term early, then modifiers, then specs.

A workable Korean title skeleton looks like: 브랜드 (brand) + 핵심 키워드 (head keyword) + 주요 속성 (key attributes) + 규격/수량 (size/quantity). Avoid the English-marketplace habit of leading with adjectives and marketing language — "Premium Ultra-Soft Luxury..." translated into Korean burns your most valuable character slots on words nobody searches.

Also resist stuffing every synonym. Coupang penalizes obviously spammy titles, and duplicated near-synonyms crowd out modifiers that would have won you a long-tail query.

• Brand first if you have any recognition; otherwise lead with the head keyword.
• Head keyword within the first ~15 characters so it survives mobile truncation.
• One head term, then two to four distinct modifiers — not four ways of saying the same thing.
• End with size, count, or color, which shoppers use as filter-style queries.

Step 3: Spacing, Hangul-vs-Latin, and the mechanics that trip up foreigners

Korean word spacing (띄어쓰기, spacing between words) affects matching. Shoppers type both 요가매트 and 요가 매트, and these are not always treated identically. The common practical approach is to include the most-searched spacing form in the title and let the other appear naturally elsewhere in your listing text.

Decide deliberately whether to write English in Latin script or transliterate into Hangul. Global brand names usually stay Latin (or appear both ways), while common nouns almost always need the Hangul form. If your brand is unknown in Korea, writing it only in Latin means shoppers who saw it once and try to recall it phonetically will never find you.

Finally, use Korean unit and format conventions. Sizes, capacities, and counts written the local way get matched to the way people search for them.

• Pick the dominant spacing variant for your head keyword; don't include both forms in one title.
• Consider Brand(Latin) 브랜드(Hangul) once each if your name is hard to recall phonetically.
• Use 개 for item counts, 세트 for sets, 인용 for person-capacity (2인용 = for two people).
• Capacity in ml/L, dimensions in cm — imperial units read as foreign and don't match queries.
• Avoid special characters and heavy punctuation; they add noise without adding matches.

Step 4: Test, then let rank data tell you which words worked

A title is a hypothesis. The only way to know whether 무선 청소기 or a different modifier earned you traffic is to track your position on each target keyword daily and change one variable at a time.

Give a title change at least a week or two before judging — reindexing plus normal ranking volatility will otherwise mislead you. Change the head keyword and a modifier simultaneously and you'll never know which one moved.

Daily rank tracking per keyword is the feedback loop that turns title writing from guesswork into iteration. SeoulRank tracks Coupang positions daily in English, which makes this practical if you can't read the results page yourself.

• Track 5–15 target keywords per SKU, mixing one head term with long-tail modifiers.
• Change one element per test cycle: head term, modifier set, or ordering.
• Watch long-tail keywords first — they move faster and signal whether your modifiers are working.
• Log every title version with a date so you can attribute rank shifts correctly.

A quick checklist before you publish

Run each title through this list. It takes two minutes and catches the majority of the damage foreign sellers do to their own listings.

• Head keyword verified against actual Korean search demand — not a translation.
• Head keyword appears in the first portion of the title, before truncation.
• No marketing adjectives occupying prime slots.
• No synonym stuffing; each word earns a distinct query.
• Units, counts, and sizes in Korean convention.
• Brand written in the form a Korean shopper could plausibly re-type.
• Target keywords logged in a tracker so you can measure the change.