Reading Coupang Reviews as a Seller: What Korean Shoppers Actually Complain About

Why competitor reviews are the cheapest research you have

Before you spend a won on ads, the reviews on your top three competitors are a free, dated, ranked list of everything Korean shoppers expect from your category — and everything the current sellers are getting wrong. Coupang surfaces reviews prominently, and 상품평 (sangpumpyeong, "product reviews") with photos carry a lot of weight in whether a shopper converts.

The trick for a non-Korean seller is that machine translation flattens the nuance. "It's so-so" in English could be 그냥 그래요 (mild shrug) or 별로예요 (actively disappointed) — very different signals. The fix is not to read everything, but to search reviews for a fixed list of Korean complaint phrases and count how often each appears. That turns a language barrier into a mechanical task.

Do this per competitor ASIN-equivalent (Coupang product page), and log the counts in a spreadsheet. Patterns show up within about 50–100 reviews.

The six complaint clusters that repeat in almost every category

Across most physical-goods categories on Coupang, negative reviews collapse into a handful of themes. Searching the review tab with Ctrl+F for these Hangul strings is faster than reading.

Cluster 1 — Delivery and packaging. Korean shoppers are conditioned by 로켓배송 (Rocket Delivery, Coupang's next-day fulfilment) to expect fast, intact arrival. Anything slower or dented gets punished hard, even when it's not your fault.

Cluster 2 — Sizing and fit. This is the single biggest killer for apparel, footwear, and anything worn or gripped. Korean sizing runs smaller than US/EU equivalents and shoppers say so bluntly.

• 배송 (delivery) / 배송이 느려요 — "delivery is slow"; 파손 (damaged), 찌그러져 (dented), 포장 (packaging)
• 사이즈 (size) / 작아요 (runs small), 커요 (runs big), 정사이즈 (true to size)
• 냄새 (smell) — plastic, chemical, or adhesive odor; a very common complaint on imported goods
• 가성비 (gaseongbi, "value for money") — used positively and negatively; 가격 대비 (for the price) signals a value judgement
• 불량 (defective), 고장 (broken), 안 켜져요 (doesn't turn on)
• 설명서 (instruction manual) / 한글 설명서 없어요 — "no Korean-language manual"

The complaints that specifically hit imported and cross-border products

Some of these clusters are disproportionately aimed at foreign sellers. If you're shipping from the US, EU, or China, assume every one of these will appear in your reviews unless you pre-empt it.

한글 설명서 없어요 ("there's no Korean manual") is the classic. So is 플러그가 안 맞아요 ("the plug doesn't fit") — Korea uses 220V / Type C-F plugs, and a US two-prong adapter shipped with an appliance generates immediate one-star reviews. For consumables and cosmetics, shoppers look for 성분 (ingredients) listed in Korean and for 정품 (authentic product) reassurance.

Also watch for 해외배송 (overseas shipping) mentioned in a negative review — it usually means the buyer didn't realise the item wasn't Rocket-fulfilled and waited two weeks. That's a listing-clarity problem, not a product problem, and it's fixable.

• Ship a one-page Korean quick-start sheet, even a printed insert — it removes an entire complaint category
• Include a 220V plug or adapter in the box if the product is electrical, and say so in the title or first image
• Show the size chart in Korean units and compare to a familiar Korean brand where possible
• If you're not on Rocket, state the delivery window in the listing above the fold

Turning complaints into listing and keyword changes

Every recurring complaint has a mirror-image search phrase. If shoppers complain 작아요 ("runs small"), then some portion of the market is actively searching for 넉넉한 핏 (roomy fit) or 오버핏 (oversized fit). If they complain about 냄새, there's demand for 무취 (odorless) or 저자극 (low-irritation).

Build the mirror list: for each complaint you counted, write the positive attribute a shopper would type into search. Then check whether those phrases actually have search volume before you rewrite your title around them — a complaint being loud in reviews doesn't guarantee it's a high-volume query. This is exactly the validation step SeoulRank is built for: you can check Korean search demand and see how the term moves in Coupang results without reading Korean.

Then put the winning attributes in three places: the product title, the first two thumbnail images, and the opening line of the detail page. Korean shoppers scan the image stack before reading anything.

• Complaint 작아요 → keyword candidates: 정사이즈, 넉넉한, 오버핏
• Complaint 냄새 → 무취, 무향, 저자극
• Complaint 불량/고장 → lean on 정품, AS 가능 (after-sales service available), 국내 배송 (domestic shipping)
• Complaint 설명서 없음 → 한글 설명서 포함 (Korean manual included) as a bullet in the detail page

Reading your own reviews: tone, response, and damage control

Korean reviewers often bury the real issue in a polite opener. A review that starts 잘 받았어요 ("received it fine") and then says 근데... ("but...") is a negative review. Train yourself to look for 근데, 다만, and 아쉬운 점은 ("the disappointing part is") — whatever follows is the actual feedback.

Respond in Korean, briefly and formally. A short reply that acknowledges the issue and states the fix (예: 한글 설명서를 동봉하도록 개선하겠습니다 — "we will improve by including a Korean manual") reads as competent. A machine-translated paragraph of English-style apology reads as a foreign seller who isn't paying attention, and other shoppers notice.

Photo reviews are the ones to triage first. They're the most visible, the most persuasive, and the most likely to be about packaging or colour mismatch — both of which you control.

• Reply within a day or two; late replies look automated
• Never argue about a size complaint — update the size chart instead and say you did
• If a complaint repeats more than three times, it's a product or packaging spec issue, not a customer issue
• Track the complaint mix monthly so you can see whether a change actually worked

A repeatable 60-minute monthly routine

You don't need a translator on retainer. You need a fixed checklist run once a month against your listing and your two closest competitors.

Copy the last ~100 reviews of each into a sheet, run a find-and-count for your Hangul complaint list, and note the top three themes per product. Then compare: the themes that appear in competitor reviews but not yours are your marketing angles; the ones in yours but not theirs are your urgent fixes.

Finish by checking whether the attribute keywords you added last month are actually moving your position for the queries you targeted. Complaint mining without rank feedback is guesswork — the loop only closes when you can see the ranking response.

• Pull 100 recent reviews per product, filtered to 1–3 stars where possible
• Count the fixed Hangul phrase list; log counts by month
• Pick one product change and one listing change per cycle — not five
• Re-check target keyword positions before the next cycle so you know which change did the work