The 5 Keyword Slots on a Coupang Listing: How to Spend Them

Why you should think in "slots," not "keywords"

Most sellers coming from Amazon or eBay treat a Coupang listing as one big text box: find good keywords, stuff them wherever they fit. That fails on Coupang for two reasons. First, the fields are weighted very differently — a term in 상품명 (product name) does not behave like the same term in the hidden 검색어 (search keywords) field. Second, Coupang actively penalizes keyword repetition and irrelevant terms in the product name, so the "spray everywhere" approach can suppress a listing that would otherwise rank.

The practical model is to treat your listing as five separate budgets. Each slot has a character limit, a ranking weight, and a job. Your head term goes in the highest-weight slot once. Everything else gets distributed so that no term is wasted and no term is repeated unnecessarily.

Before you allocate anything, you need a ranked list of Korean search terms with real volume behind them — not Google Translate output. A translated phrase is frequently a term Korean shoppers never type.

• Slot 1 — 상품명 (product name): highest weight, most constrained, most heavily policed.
• Slot 2 — 검색어 / 검색 태그 (search keywords, buyer-invisible): your synonym and misspelling dump.
• Slot 3 — 카테고리 + 상품 속성 (category and attribute/spec fields): filter eligibility, indirect but decisive.
• Slot 4 — 옵션명 (option/variant names): overlooked, searchable, and shown in results.
• Slot 5 — 상세페이지 (detail page) text and image alt content: weak for search, strong for conversion.

Slot 1 — 상품명 (product name): one head term, zero repetition

Coupang publishes a recommended product name structure and it is worth following literally: 브랜드 (brand) + 제품명 (product name) + 상품군 (product type) + 색상/사이즈/수량 (color/size/quantity). Listings that deviate — all-caps marketing copy, bracketed promos like [무료배송] (free shipping), seller-name prefixes — tend to get filtered out of search or flagged in quality checks.

Keep it short. Korean product names convey more per character than English, and longer names dilute relevance per term rather than adding it. A tight name in the 30–50 character range generally outperforms a 100-character keyword chain. Use your single highest-volume head term here, exactly once, in the form shoppers actually type.

Two Korean-specific traps. 띄어쓰기 (word spacing): Korean shoppers type both 무선이어폰 and 무선 이어폰 (wireless earbuds), and these can behave as different queries — pick the higher-volume spacing for the name and put the other variant in Slot 2. And loanword script: 텀블러 vs "tumbler" vs 보온병 (vacuum flask) are three distinct queries, not three spellings of one.

• Follow Coupang's brand + product + type + variant order; don't reinvent it.
• Use each keyword once. Repetition is a demotion risk, not a boost.
• Strip promotional words: 할인, 무료배송, 최저가, 1위 — these are commonly restricted in names.
• Decide spacing by volume, not by grammar. Put the loser in the tag field.
• If you have a real Korean-recognized brand, lead with it; if you're unknown, lead with the category term instead.

Slot 2 — 검색어 (search keywords): the synonym and misspelling budget

This is the buyer-invisible tag field in Wing (Coupang's seller back office). It carries less weight than the product name, but it's the only place you can safely cover variation without polluting your title. Treat it as a budget of distinct concepts, not a word count to max out.

Spend it on: spacing variants, Hangul/English script variants, regional or generic synonyms, common 오타 (typos) that Korean keyboards produce, use-case phrases (선물용, 자취생, 캠핑용 — gift, single-household, camping), and seasonal modifiers. Do not spend it on: words already in your product name, competitor brand names (a takedown risk), or terms unrelated to the product.

This is also where English-speaking sellers lose the most ground, because the variants are invisible unless you can read Korean search data. Tools like SeoulRank surface Korean query variants and their relative volume in English, which is usually enough to fill this field properly in one sitting instead of guessing.

• One concept per tag; don't repeat the head term in five slightly different forms if only one is searched.
• Include at least one 조사-free noun chunk — Korean shoppers search in bare nouns, not sentences.
• Add two or three use-case/occasion terms; they convert well and face less competition.
• Skip competitor brand names entirely — it's one of the faster ways to get a listing pulled.
• Revisit the field quarterly; Korean trend terms turn over fast.

Slot 3 — 카테고리 and 상품 속성 (category and attributes): keywords you don't type

Korean shoppers filter heavily. On a category page, the left-rail filters — 색상, 용량, 소재, 연령, 호환기종 (color, capacity, material, age, compatible model) — are how people narrow 4,000 results to 40. If an attribute field is blank, your product simply does not exist for that filter, no matter how good your title is.

Category choice is the bigger lever. Coupang's ranking is largely within-category, so a misfiled product competes in the wrong pool and inherits the wrong filter set. Check where the top-selling equivalents of your product actually sit before you pick — the obvious category is not always the one the market uses.

Fill every optional attribute, even the ones that feel pointless. The cost is ten minutes; the benefit is eligibility for filter traffic you'd otherwise never see, plus better matching on long-tail queries that include the attribute value.

• Mirror the category of the top 3–5 organic results for your head term.
• Fill optional attributes — blank fields silently remove you from filtered views.
• Compatibility fields (phone model, car model, appliance model) are high-intent and often sparsely filled by competitors.
• If your product spans two plausible categories, test one, measure, then move — don't split inventory across duplicate listings.

Slot 4 — 옵션명 (option names): the slot almost nobody optimizes

Option names appear in search results, in the buy box, and in the item page URL structure. They're also indexed, which means a variant named "02" wastes a searchable string while "블랙 / 500ml" (black / 500ml) captures color and capacity queries for free.

Write options the way a Korean shopper would describe the thing: concrete attribute values, consistent order, no internal SKU codes. If you sell a 3-pack and a 5-pack, say 3개입 / 5개입 (3-piece / 5-piece) rather than "Set A / Set B." This also reduces returns, which matters because Coupang's ranking is sensitive to post-purchase signals.

One structural decision worth getting right early: group genuine variants of one product under a single listing rather than publishing them separately. Consolidated sales history and reviews compound; split listings compete with each other and dilute both.

• Use real attribute values — color, size, capacity, count — never SKU codes or "Type 1."
• Keep attribute order identical across every option for scannability.
• Put the variant you most want to sell first; it's the default shown.
• Don't duplicate your full product name inside each option — it reads as spam and adds nothing.

Slot 5 — 상세페이지 (detail page): weak for ranking, decisive for conversion

Korean detail pages are almost entirely images, which means most of your copy isn't crawlable text. Don't plan your keyword strategy around it. Plan your conversion rate around it — and on Coupang, conversion rate feeds back into rank, so this slot earns its keep indirectly.

What Korean buyers look for in the detail page, roughly in order: clear 사이즈/스펙 (size and spec) tables, real-use photography with a sense of scale, material and origin disclosure, 교환/반품 (exchange and return) terms, and any required certifications such as KC 인증 for applicable categories. Machine-translated Korean is immediately obvious and costs you trust at exactly the wrong moment.

Where you do have text fields — bullet summaries, A/S (after-service) notes, any plain-text sections your template supports — use natural sentences containing your secondary terms. Write for a human first; the keyword value here is a rounding error compared to the conversion value.

• Get the Korean copy reviewed by a native speaker, even if it's only 200 words.
• Lead the page with a spec table — Korean shoppers scroll to it first.
• State shipping origin and lead time plainly if you ship cross-border; ambiguity kills conversion.
• Include certification marks where your category requires them; missing them can get a listing suspended.

Putting it together: a one-hour allocation pass

Work in order, top-weight slot first, and never let a term appear in two slots unless it's genuinely a different query form. Start with a ranked list of 15–25 Korean terms for your product, grouped into one head term, three to five mid-tail terms, and the rest long-tail or variant.

Then measure. The only reliable read on whether your allocation worked is daily rank for your target terms, tracked over a few weeks — Coupang rankings move, and a single-day check tells you nothing. Change one slot at a time so you can attribute the movement. SeoulRank handles the daily tracking side in English if you'd rather not check positions by hand in Korean.

Expect to revisit Slot 2 and Slot 4 every quarter, Slot 1 only when the data clearly says your head term is wrong, and Slot 3 whenever Coupang adds filters to your category.

• Head term → product name, once, in the highest-volume spacing.
• Variants, typos, synonyms, use-cases → search keyword field.
• Every attribute value → category and attribute fields, no blanks.
• Color/size/count → option names, in plain Korean.
• Everything else → detail page, written for humans by a human.