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

Why a Translated Title Almost Never Ranks

A translation answers the question "what is this product called in Korean?" A localized title answers a different question: "what words does a Korean shopper type into the Coupang search bar when they want this product?" Those two answers overlap less than most sellers expect.

Coupang's search engine matches query tokens against your 상품명 (product name), plus category, attributes, and tags. If your title contains the dictionary-correct Korean word but shoppers search a loanword, a brand-as-category term, or an abbreviation, you simply do not appear in the result set — no amount of ad spend fixes a missing token.

The practical consequence: title writing is a keyword research task first and a copywriting task second. Write the keyword set, then assemble the sentence.

• Translation optimizes for correctness; localization optimizes for retrievability.
• A missing token usually means zero impressions, not fewer impressions.
• Korean shoppers type short, noun-stacked queries — rarely full phrases.

Step 1: Build the Real Query Set Before You Write Anything

Start by collecting every way Korean buyers might name your product. There are usually three or four competing names for the same item, and the winner is rarely the one a translator picks.

Pull candidates from Coupang's own autocomplete (type the first two or three syllables and record the suggestions), from the "연관검색어" (related search terms) strip, and from the titles of the top 20 organic listings in your category. Those top listings are a free, continuously A/B-tested dataset of what works.

Then you need volume and competition data to choose between candidates. This is where English-language tooling helps — SeoulRank shows Korean search volume and difficulty with English glosses, so you can compare 무선이어폰 and 블루투스이어폰 without guessing which one shoppers actually use.

• Record autocomplete suggestions for every root term — autocomplete reflects real query frequency.
• Copy the top 20 competitor titles into a sheet and count token frequency; repeated tokens are your required vocabulary.
• Note which tokens appear in the first 15 characters of winning titles — that position is not accidental.
• Check both the Sino-Korean/native term and the English loanword version of every noun.

Step 2: Loanword vs. Native Korean — Pick the One Shoppers Type

Korean has a huge inventory of English loanwords written in Hangul, and for many consumer categories the loanword outranks the native term in search. 텀블러 (tumbler) is how people search for insulated bottles; the more native-sounding alternatives feel formal and get typed far less. Meanwhile, in food, home, and traditional categories, the native word usually dominates.

There is no rule you can apply blindly — it is category by category, and it shifts over time. The safe method: check both spellings in autocomplete and in the top listings, and if both have real volume, include both in the title as long as it still reads naturally.

Also watch romanization variants. Korean renders the same English word multiple ways (e.g., different Hangul spellings for "gel," "vitamin," or brand names). Coupang does not always treat them as synonyms, so the dominant spelling matters.

• Loanwords tend to win in beauty, electronics, fitness, and fashion.
• Native Korean terms tend to win in food, kitchenware, bedding, and health staples.
• When two spellings both have volume, include both — but never three variants of the same word.
• Check the dominant Hangul spelling of your own brand name; buyers will search it that way, not in Latin letters.

Step 3: Order and Spacing Are Ranking Decisions

Korean product titles follow a conventional stack: 브랜드 (brand) → 핵심 키워드 (core keyword) → 주요 속성 (key attributes: size, capacity, color, count) → 용도/대상 (use case or target user). Deviating from this reads as amateur to shoppers and buries your core keyword below the fold on mobile, where most Coupang traffic lives.

Spacing (띄어쓰기, word spacing) matters more than it looks. Korean compounds can be written joined or spaced, and shoppers type them both ways. Where a term is commonly written joined in search (like 무선청소기, cordless vacuum), write it joined; adding a space can weaken the match for the joined query.

Mobile truncation is the practical constraint. Assume only the first portion of your title is visible in the search grid, and make sure the brand plus the primary keyword plus the single most decisive spec all land there.

• Front-load: brand, core noun, then the spec that drives the purchase decision (capacity, size, pack count).
• Match the spacing convention used by the top-ranked listings, not grammar textbooks.
• Avoid marketing adjectives in the first half — "premium," "best," "hot deal" consume prime characters and match nothing.
• Do not repeat the same keyword; repetition does not boost rank and can trigger policy flags.
• Keep special characters and emoji out — Coupang restricts them and they break tokenization.

Step 4: Localize the Specs, Not Just the Nouns

Korean shoppers search by units and formats that may differ from your home market. Sizes, counts, and measurements are often part of the query itself, so writing them in the local convention is a retrieval issue, not a formatting preference.

Common conversions: apparel sizing to Korean numeric sizes alongside S/M/L, shoe sizes in millimeters (250, 260, 270), weight and volume in metric with the unit written in Latin characters (ml, g, kg) as Koreans do, and pack counts using the counter 개 or 입 (e.g., 30개입 for a 30-count pack).

Also localize the use-case token. Terms like 선물용 (for gifting), 캠핑용 (for camping), 자취 (living alone / small household), or 유아용 (for infants) are high-intent modifiers that Korean shoppers genuinely type, and they often carry far less competition than the head keyword.

• Shoe sizes in mm; apparel with both Korean numeric and S/M/L.
• Pack counts with the correct Korean counter (개입, 매, 포, 세트).
• Keep metric units in Latin script (500ml, 1kg) — that is the search convention.
• Add one high-intent use-case modifier such as 선물용 or 자취 if it fits naturally.

Step 5: Test, Track, and Iterate

A title is a hypothesis. The only way to know whether 무선이어폰 or 블루투스이어폰 serves you better is to publish, track daily rank for both terms, and watch which one moves.

Change one variable at a time. Swap the core keyword, hold everything else, and give it enough days to stabilize — Coupang rank is noisy day to day, and reacting to a single day's drop is how sellers destroy a title that was working. Daily rank tracking across your target keyword set, which is what SeoulRank was built for, makes the signal readable without needing Korean.

Build a small rhythm: audit titles quarterly, re-pull autocomplete (query language drifts, especially in beauty and electronics), and check whether new competitors have introduced a keyword you are missing.

• Track 5–15 target keywords per listing, including the ones you chose not to use.
• Change one token per test cycle; log the date of every title edit.
• Watch impressions and click-through, not just rank — a keyword that ranks but does not convert is the wrong keyword.
• Re-audit after any category or seasonal shift; search vocabulary moves faster than product catalogs do.