Rocket Growth vs Marketplace: The Cross-Border Seller's Decision

First, get the three Coupang models straight

Most confusion for foreign sellers comes from lumping everything with a rocket icon together. Coupang actually runs three distinct arrangements, and only two of them are decisions you make as a third-party seller.

Rocket Delivery (로켓배송, "Rocket Delivery") is a wholesale relationship: Coupang buys your inventory, owns it, prices it, and sells it. You are a vendor, not a seller. You do not control the listing or the price, and you generally need Coupang Retail to invite you.

The two you actually choose between are Rocket Growth and Marketplace.

• 로켓그로스 (Rocket Growth) — you own the inventory, but store it in Coupang's fulfillment centers. Coupang picks, packs, ships, and handles returns and customer service on the delivery side. Functionally Korea's closest analogue to FBA. It was previously branded 제트배송 (Jet Shipping), so older guides and forum posts may use that name.
• 마켓플레이스 (Marketplace), also called 판매자배송 (seller delivery) — you list, you hold stock wherever you like, you ship, you handle returns. Full control, full operational burden.
• A single seller account can run both simultaneously, SKU by SKU. This is the point most cross-border sellers miss: it is not an either/or company-level decision.

The visibility gap is the real variable, not the fee

Korean shoppers are trained to filter for rocket-badged items. The Rocket filter is one of the most-used facets on Coupang search, and next-day (often same-day in Seoul) delivery is treated as a baseline expectation rather than a premium. A Marketplace listing shipping from Shenzhen or New Jersey with a 7–14 day window is competing against that expectation on the same results page.

In practice this shows up two ways: lower click-through from search, and a much lower conversion rate on the product page for identical products at identical prices. Because Coupang's ranking system rewards sales velocity and conversion, weak conversion compounds into weaker organic position over time.

This is where rank tracking earns its keep. If you're running the same SKU both ways, or migrating a SKU from Marketplace to Rocket Growth, watching daily keyword position before and after the switch tells you what the badge is actually worth in your category — some categories are far more rocket-dominated than others. SeoulRank tracks Coupang positions daily in English, which makes that before/after comparison readable without a Korean-speaking VA.

• Check the top 20 results for your three main keywords and count how many carry a rocket badge. If it's 15+, Marketplace-only is a hard road in that category.
• Categories where Marketplace still competes well: bulky/oversized items, made-to-order and personalized goods, high-ASP niche products, B2B-ish supplies, and anything where the buyer expects a lead time anyway.
• Categories where Rocket Growth is close to mandatory: consumables, beauty, small home goods, pet supplies, phone accessories, kitchen — anything cheap, repeat-purchase, and heavily comparison-shopped.

How the cost math actually works

Rocket Growth charges you on a per-unit basis for fulfillment plus storage over time, and the sales commission (판매수수료, selling commission) applies in both models by category. So the incremental cost of Rocket Growth versus Marketplace is fulfillment + storage + inbound freight into Korea, minus whatever you were paying to ship cross-border per order.

For most small parcels this comparison is not close. Direct cross-border shipping of a single low-value unit into Korea is expensive per unit and slow. Consolidating a pallet or container into a Coupang fulfillment center and paying domestic per-unit fulfillment usually wins decisively on unit economics — the catch is that you pay for the whole shipment up front and you're exposed if it doesn't sell.

Build the model per SKU, not per catalogue. The variables that flip the answer are unit weight, unit price, and sell-through rate.

• Model the landed cost per unit for a full inbound shipment, including customs duty, VAT (recoverable if you're registered), and inland trucking to the fulfillment center.
• Add a realistic storage cost assumption based on your expected days-of-cover, not your best case. Slow SKUs quietly eat margin in storage.
• Subtract what you currently spend per order on cross-border shipping and the labour cost of handling Korean-language returns yourself.
• Add a returns allowance. Korean marketplace return rates in apparel and fashion run materially higher than in home or consumables — plan for it rather than being surprised.
• Sanity-check against your competitors' listed prices. If the rocket-badged incumbents are priced below your Rocket Growth landed cost plus commission, the SKU is not viable in that category at all.

The compliance and setup work is not symmetrical

Marketplace lets you start selling with a lighter setup. Rocket Growth requires you to actually import goods into Korea, which means a customs clearance identity, product-level compliance for the category, and Korean labelling on the physical unit.

For non-Korean entities this is the step that stalls launches. You typically need a local presence to act as importer of record — either a Korean entity, a partner, or a service provider that performs that function. Certain categories add certification requirements: KC certification (KC 인증, Korea Certification mark) for electrical, children's, and various consumer goods; food and cosmetics fall under MFDS oversight with their own import declaration process.

None of this disappears in Marketplace — it just gets deferred, because goods clear as low-value personal-use imports per order rather than as a commercial bulk import. That deferral is convenient early and becomes a liability if you scale, because compliance scrutiny does not stay proportional to your intent.

• Confirm category certification requirements before you book freight, not after. Goods held at customs for missing KC marks are an expensive lesson.
• Korean-language labelling (product name, importer, country of origin, ingredients or specs, safety warnings) generally needs to be on the unit or an applied sticker before it reaches the fulfillment center.
• Barcoding and inbound packing requirements are strict. Rejected inbound shipments cost you weeks.

A practical decision path

• Test on Marketplace first, at small scale, on 5–15 SKUs. Accept that conversion will be depressed by the delivery time — you're buying demand data, not profit.
• Watch which SKUs still sell despite the handicap. Those are your genuine demand signals, and they're the ones worth importing.
• Track keyword rank and search volume in parallel. A SKU that ranks page 3 on its main keyword may be underselling because of position, not product-market fit. Distinguish the two before you commit inventory.
• Send a deliberately small first inbound shipment — enough for roughly 45–60 days of your Marketplace run-rate, adjusted upward for the conversion lift you expect from the badge. Undershooting costs you a stockout; overshooting costs you cash and storage.
• Once Rocket Growth is live, keep the slow tail, oversized items, and any variant with unstable demand on Marketplace. A hybrid catalogue is normal and efficient.

What changes after you switch

Expect the lift to show up in conversion rate first and organic rank second, with a lag. Ranking improvements follow accumulated sales velocity, so give a migrated SKU several weeks before judging it. Sellers who switch, see flat rank for ten days, and panic into heavy ad spend usually just buy traffic they were about to get for free.

You also lose some control. Coupang handles the customer touchpoint on delivery and returns, which is a relief operationally but means service complaints attach to your listing without you being able to intervene directly. Product quality and packaging durability matter more, not less.

Finally, watch your inventory health metrics. Rocket Growth rewards sellers who keep stock available and turning; chronic stockouts undermine the ranking gains you paid to get. Set reorder points off your Korean sell-through rate plus realistic ocean freight and customs lead time — for most cross-border sellers that's a lead time measured in weeks, not days.

• Re-baseline your keyword positions the week before migration so you have a clean comparison.
• Rebuild your ad bids after the switch — a rocket-badged listing converts better, so the same keyword can support a higher bid profitably.
• Treat the first inbound as a calibration run. Your second order quantity is the one that matters.