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

Korean shoppers read prices in 만 (10,000) units, not thousands

Western pricing instinct is built around the dollar and round decades: $19.99, $49, $99. Korean price perception is built around 만원 (man-won, 10,000 KRW). Shoppers mentally bucket products as "1만원대" (in the 10,000s), "2만원대" (20,000s), and so on. Crossing from 19,900원 to 20,900원 is not a 5% increase in the shopper's head — it moves your product into a different mental category.

This is the single most important pricing rule on Coupang. Before optimizing anything else, find which 만원대 band your product belongs to based on competitors, then price at the top of the band you want to be perceived in, not the bottom of the next one.

Practical read: 9,900원, 19,900원, 29,900원, 39,900원 and 49,900원 are the anchor points of Korean e-commerce. They appear so often that shoppers treat them as "normal price" rather than as a discount trick.

• Never price at 20,000원, 30,000원 or 50,000원 exactly — the psychological penalty is real and you gain nothing.
• If your cost structure forces you above a band, go clearly above it (e.g. 23,900원) rather than barely above (20,500원), which reads as greedy.
• For premium or gift positioning, round numbers in the higher bands (e.g. 89,000원) can signal confidence — this is the one exception.
• Check whether your category clusters at X9,900 or X8,000 — beauty and food often use 9,900; industrial and B2B-ish goods use flatter numbers.

The 900-won ending and why 990 is not the same as 999

Korean retail settled on 900 and 90 endings, not 99. You will see 12,900원 and 4,990원 constantly; you will rarely see 12,999원. A price ending in 999 looks slightly foreign or machine-generated to a Korean shopper, and machine-generated is exactly the vibe cross-border sellers need to avoid.

Below 10,000원, the common endings are 900 and 990 (3,900원, 5,900원, 6,990원). Above 100,000원, endings loosen: 129,000원 or 149,000원 are typical, and the trailing 900 matters much less because the leading digits dominate perception.

• Under 10,000원: use X,900 or X,990.
• 10,000–99,000원: use XX,900.
• Over 100,000원: use XXX,000 — trailing 900s start to look fussy.
• If you sell multipacks, keep the ending consistent across variants so the option list looks deliberate (9,900 / 18,900 / 26,900), not random.

Free shipping thresholds shape the price, not the other way around

Korean shoppers are unusually sensitive to shipping fees — partly because 로켓배송 (Rocket delivery) and Coupang's Wow membership have trained the market to expect fast, free, no-surprise delivery. A 9,900원 item with a 3,000원 shipping fee loses to a 12,900원 item with free shipping almost every time, even though the buyer pays more.

So work backwards. Decide the all-in price the shopper sees, then fit your margin inside it. If you are on 마켓플레이스 (third-party marketplace) fulfillment rather than Rocket, your price has to absorb shipping or you need to hit whatever free-shipping threshold you advertise.

Bundling is the standard Korean answer to a thin-margin single unit. Rather than sell one unit at 6,900원 plus shipping, sell a 3-pack at 17,900원 free-shipped. The unit price is barely lower but the transaction is far more attractive and your logistics cost per order drops.

• Always compare your price to competitors on a delivered-price basis, including their shipping fee — that is what the shopper compares.
• Design a bundle that lands just above your free-shipping threshold rather than just below it.
• If a competitor is Rocket and you are not, assume you need a visible price advantage or a clearly bigger pack size to win the click.
• Watch for 묶음배송 (combined shipping) expectations — shoppers buying two of your listings expect one shipping fee.

Discount framing: what 할인 actually needs to look like

Coupang shows a struck-through original price and a percentage off. Korean shoppers are experienced enough to discount the discount — a permanent "70% 할인" (70% off) on an unknown foreign brand reads as fake, not as a bargain. Credible framing beats aggressive framing.

The percentages that feel real in Korea tend to be modest and specific: 10%, 15%, 23%. A discount that leaves the final price on a clean psychological anchor (from 24,900원 to 19,900원) does double duty — it looks like a real markdown and lands in a lower 만원대 band.

Also remember Coupang's own promotional mechanics and coupon stacking. If a shopper can apply a card or Wow discount on top of yours, your displayed price may not be the price they mentally record. Model the worst case before you commit to a coupon.

• Aim discounts so the final price lands on X9,900 — the markdown and the anchor reinforce each other.
• Avoid perpetual deep discounts; they compress your reference price permanently and make future full-price sales impossible.
• Use 1+1 or 2+1 pack framing instead of percentage cuts when you want volume without resetting the reference price.
• For a launch, a small discount plus a strong review-gathering push usually outperforms a big discount with no social proof.

Price is a ranking input, so test it like one

On Coupang, price feeds conversion, conversion feeds ranking, and ranking feeds volume. That loop means a price test is never just a margin test — it can move your position on the keyword pages you care about. Change price and watch rank, not only revenue.

The clean way to run this: pick two or three candidate price points inside the same 만원대 band, hold everything else constant (images, title, coupons), run each for a full week including a weekend, and log your daily rank on your top 10–20 keywords alongside units sold. Weekly cycles matter because Korean shopping traffic skews heavily to evenings and Sundays.

If you don't read Korean, the hard part is knowing which keywords to watch and what your competitors' prices are on those exact terms. Tools like SeoulRank exist for this — Korean keyword data and daily Coupang rank tracking presented in English — but even a manual spreadsheet of ten keywords checked daily beats guessing.

One caution: do not change price and creative in the same week. You will not know which one moved the needle, and on Coupang the recovery time for a bad price move is measured in weeks, not days.

• Test within a band first (18,900 vs 19,900), then across bands (19,900 vs 23,900) — cross-band tests are strategic, not tactical.
• Log daily rank for your money keywords before, during and after the change; conversion effects show up in rank within days.
• Never test during a major sales event — the traffic mix is abnormal and the result won't generalize.
• Track units and margin per day, not revenue; a 10% price cut that lifts units 12% can still lose money.
• Keep a written price history. When rank drops six weeks later, you will want to know what you changed.

Common cross-border pricing mistakes

Most foreign sellers on Coupang lose on price for boring, fixable reasons: they convert from USD and land on ugly numbers, they underestimate delivered-price comparison, or they price like a premium brand without any of the Korean-language trust signals that justify a premium.

Fix the mechanics first. A converted price of 21,347원 signals "imported listing, nobody is paying attention" before the shopper reads a single word of your description.

• FX-converted prices: always round to a Korean anchor after converting, and re-check when the won moves.
• Pricing above the category leader with no reviews, no 로켓배송, and a machine-translated title — the premium has to be earned.
• Ignoring option pricing: Coupang shows the lowest option price in search, so a cheap sample-size option can help clicks but hurt perceived value if mismanaged.
• Forgetting VAT and Coupang commission when setting the anchor — build them into the band decision, not as an afterthought.
• Setting one price for all of Korea's seasons; many categories (fans, heaters, gift sets) have sharp seasonal price ceilings.