Korean Price Psychology: How to Price a Product for Coupang Shoppers

Why Korean price perception is not just "charm pricing" translated

Korean prices are quoted in 원 (won), with no decimals and thousands separators everywhere. That single fact changes how shoppers scan a listing. A US buyer parses $19.99 as "about twenty dollars." A Korean buyer parses 19,900원 as "a nineteen-thousand-won item." The leading digit and the digit count carry almost all the perceived weight, and everything after the comma is treated as rounding noise.

This is why the dominant tactic on Coupang is not shaving one unit off the price — it is shaving one unit off the leading digit while keeping the number of digits the same. 19,900원 and 20,000원 differ by 0.5%, but they sit in different mental buckets. Sellers who mechanically convert $20 to roughly 27,000원 and publish that number are leaving conversion on the table for no margin gain.

Also note that 만원 (man-won, ten thousand won) is the everyday counting unit. Shoppers think in 1만원, 3만원, 5만원 chunks. Crossing a 만원 boundary is a real psychological event; moving within one is not.

The number patterns Coupang shoppers actually expect

Scan any Coupang category and you will see the same endings again and again. These are not arbitrary — they are the conventions Korean shoppers have been trained on for two decades of online retail, and a price that ignores them reads as either careless or foreign.

Use these as defaults, then deviate only when you have a reason:

• **-900 endings** are the workhorse: 9,900 / 14,900 / 19,900 / 29,900. This is the single most common pattern in consumer goods.
• **-800 or -700 endings** appear when a seller wants to undercut a -900 competitor without dropping a full tier: 19,800원 against a rival's 19,900원.
• **-90 endings on cheap items**: 3,490원, 5,990원 for consumables and small accessories.
• **Round 만원 numbers** signal premium or bundled value: 50,000원, 100,000원. Beauty, health supplements, and gift sets often use clean rounds deliberately.
• **Avoid 4 where you can** — 4 (사) is homophonous with 死 (death). It is not a hard taboo in pricing and you will see 4-prices constantly, but for gift categories, weddings, and anything given to elders, a 44,000원 price point is a needless risk. 43,900원 or 45,000원 costs you nothing.

Anchoring, 할인율, and the discount-rate reflex

Coupang surfaces a 정가 (jeongga, list price) with a strikethrough and a 할인율 (harinyul, discount rate) badge in percent. Korean shoppers are extremely fluent at reading that badge and moderately skeptical of it. A permanent "70% off" on an unknown brand reads as fake; a 15–30% discount on a plausible list price reads as a real deal.

The practical move is to set your list price at a number that is defensible — matching your own brand site, your Naver Smart Store listing, or your export price sheet — and then discount into a psychologically clean sale price. Anchor 39,000원, sell at 29,900원. The shopper sees a 23% badge and a 2만원대 (twenty-thousand-won range) price. Both signals fire.

One caution: Coupang and Korean fair-trade rules take inflated reference prices seriously, and inconsistent list prices across channels invite complaints. Keep your 정가 the same everywhere.

• Set the sale price first (the number you want shoppers to *feel*), then back into a list price that produces a believable percentage.
• Discount rates in the 20–40% band tend to read as credible promotions rather than clearance.
• If you run Coupang's own promotions, model the stacked final price — a coupon can drag you across a 만원 boundary you did not intend to cross, in either direction.

Shipping, Rocket delivery, and the true comparison price

Korean shoppers compare *delivered* price, and they do it fast. 무료배송 (mu-ryo-baesong, free shipping) is close to table stakes in most categories. A 15,000원 item with 3,000원 shipping loses to a 17,900원 item with free shipping, even though the buyer pays more, because the first listing forces a calculation and the second does not.

로켓배송 (Rocket Delivery) changes the math again. Shoppers will pay a visible premium — often several thousand won — for next-day arrival, and Rocket items dominate the default sort for many queries. If you are on 3PL/marketplace fulfillment competing against Rocket sellers, you are not competing on the same price ladder. Either match the Rocket price and accept thinner margin, or price clearly below it and make the value trade explicit in your listing images.

Bundle sizing is the standard escape hatch. A 2-pack or 3-pack at 24,900원 is not directly comparable to a competitor's single unit at 12,900원, which is exactly the point.

• Build shipping into the item price and show 무료배송 rather than splitting it out.
• Check the delivered price of the top 5 organic results for your main keyword — that is your real price ceiling.
• If you cannot win on unit price, change the unit: multipacks, refill sets, or a bundled accessory.

Finding your price band before you launch

Pricing blind is the most expensive mistake cross-border sellers make on Coupang. The distribution of prices for a Korean search term is often bimodal — a cluster of cheap domestic/Chinese-sourced listings and a smaller cluster of branded items at 2–3x — and you need to know which cluster you are joining before you set a number.

Do the research at the keyword level, not the product level. Search the actual Korean term shoppers use (not the English product name), record the visible prices in the first two pages of results, note which are 로켓배송, and find the gaps. Frequently you will find a hole at, say, 2만원대 where everything is either 12,900원 or 35,000원. That gap is a position.

This is also where a tool helps if you do not read Korean. SeoulRank surfaces the Korean keywords for your category in English and tracks where listings rank day to day, so you can see whether a competitor's price cut actually moved them up — which tells you far more than a one-time price scrape.

• Log the top 20 prices for your main keyword in a sheet, tagged by fulfillment type and review count.
• Ignore listings with near-zero reviews when setting your ceiling; they are not proven demand.
• Note the *modal* price, not the average — averages get dragged around by outlier bundles.

Testing and adjusting without wrecking your listing

Coupang rewards sales velocity and conversion. A price change that lifts conversion can compound into better placement, which lifts sales again — and the reverse is equally true, so avoid thrashing. Change one variable, hold it for at least a week or two, and watch rank as well as units.

The cleanest tests are tier-boundary tests: 21,900원 versus 19,900원, or 34,900원 versus 29,900원. These cross a mental threshold, so if the price point matters at all, you will see it. Testing 19,900원 against 19,500원 will teach you nothing.

Finally, watch your FX exposure. If you price in won off a spot rate and the 원 moves against you, a 3% currency swing can eat a thin margin. Set your won price with a buffer and revisit quarterly rather than reacting to every wobble — frequent repricing confuses returning shoppers and can trip Coupang's own price-history displays.

• Test across a 만원 boundary, not within one.
• Hold each price for a full week minimum so weekday/weekend mix evens out.
• Track rank alongside units — a price cut that raises units but not rank may just be cannibalizing margin.
• Rebuild your won price from landed cost plus FX buffer, not by converting your home-market retail price.