Coupang vs Amazon: Korea’s Logistics Battle Explained

Coupang delivered 99.6% of South Korean orders within 24 hours in 2025. Amazon covered 98% of Japan’s population with next-day delivery that same year. Two companies, both obsessed with logistics-first ecommerce, both generating more than $30 billion in annual revenue from their core Asian markets^[1] — and yet they’ve arrived at that number through radically different playbooks. Coupang built a vertically integrated delivery empire across a single 51-million-person country,^[2] posting $34.5 billion in net revenues in 2025, up 14% (18% in constant currency),^[3] with 99% of it from South Korea (Coupang FY2025 Earnings). Amazon spread its Asian operations across Japan, India, Australia, and Singapore (where it has closed its local fulfilment operations, with amazon.sg continuing as a cross-border store), leaning on its global marketplace model and third-party seller ecosystem.

The contrast matters because it exposes a foundational question about ecommerce in Asia: does the company that owns the last mile win, or the company that owns the platform?

How big is each company in Asia?

Start with Coupang. Net revenues hit $34.5 billion in 2025, up 14%,^[3] driven overwhelmingly by the South Korean market where it commands a 25% share of ecommerce by revenue and roughly 40% of online retail traffic (Morningstar, TMO Group). South Korea’s ecommerce market reached $230 billion in 2024^[4] and is projected to hit $336 billion by 2027, growing at 13% annually (WPIC). Coupang is gaining share faster than the market is growing — Euromonitor projects 4% CAGR for the overall market against 10% for Coupang through 2028. Active customers exceeded 22 million in 2025, meaning nearly half of all South Korean adults buy from Coupang regularly.

Amazon’s Asian footprint is broader but thinner. Amazon Japan net sales were $30.7 billion in FY2025 (up from $27.4 billion in FY2024),^[1] making it Amazon’s fourth-largest market globally^[1] and about 4.3% of Amazon’s $716.9 billion group net sales^[1] (Amazon 10-K). India sits behind that, with Amazon holding 30-35%^[5] of a fast-growing ecommerce market where it announced in December 2025 that it will invest more than $35 billion through 2030, on top of roughly $40 billion invested since 2010^[6] (New Straits Times). Singapore served as the Southeast Asian gateway through amazon.sg, drawing 7.3 million monthly visitors but barely registering against Shopee and Lazada‘s dominance in the region. Amazon closed Amazon Fresh and its local fulfilment operations in Singapore on 6 July 2026^[7]; amazon.sg continues as a cross-border store (US, Japan and Germany catalogues)^[8] (The Next Web).

So Coupang owns one market deeply. Amazon touches several markets lightly. That distinction explains almost everything that follows.

What’s the fundamental difference in their Asian strategies?

Coupang’s model is infrastructure-first, marketplace-second. The company operates over 100 fulfilment centres across South Korea, covering 5.9 million square metres of warehouse space, with 70% of the population living within ten minutes of one^[9] (Coupang SEC Filing). Its Rocket Delivery service offers Dawn Delivery — order by midnight, receive before 7am^[9] — and same-day options for millions of SKUs, including fresh groceries. Coupang employs its own delivery drivers, controls its own fleet, and manages the entire chain from warehouse shelf to customer doorstep. It’s not a marketplace with logistics bolted on; it’s a logistics company with a marketplace bolted on.

Amazon’s Asian approach is the inverse. In Japan, it operates 25-plus distribution centres and 50-plus delivery stations, but relies heavily on third-party sellers and contracted delivery partners. In India, it runs a hybrid model — mixing Fulfilment by Amazon warehouses with seller-fulfilled orders — while competing against the Walmart-backed Flipkart and the explosive growth of quick commerce players like Blinkit and Swiggy Instamart. Amazon’s strength is the platform: Prime membership, the trusted brand, AWS cross-subsidisation, and global seller tools that make it easy for merchants to list across 20-plus country sites simultaneously.^[10]

The philosophical gap is real. Coupang spent years burning cash to build physical infrastructure before turning profitable. Amazon Asia uses infrastructure that already exists globally, optimising margins from day one in each new market.

Where has Amazon succeeded in Asia, and where has it failed?

Japan is the clear success story. Amazon.co.jp is the country’s largest ecommerce platform by transaction volume, with nearly 600 million monthly visitors^[11] and next-day delivery covering all 47 prefectures.^[12] Japan’s $258 billion ecommerce market is growing at 7.6% annually^[13] (BusinessWire), and Amazon has steadily gained ground against Rakuten by investing in convenience-store pickup points, same-day delivery, and aggressive Prime pricing. It works in Japan because the market rewards reliability, breadth of selection, and a premium brand — all things Amazon does well.

India is a work in progress. Amazon holds 30-35% market share against Flipkart’s 48%,^[5] but the $40 billion cumulative investment^[6] signals long-term conviction (Kotak Securities). During Prime Day 2025, Amazon India processed 18,000 orders per minute, a 50% jump year-on-year.^[14] The company launched Amazon Now for quick commerce, with orders growing 25% month-on-month.^[15] It’s competitive but not dominant, and Meesho’s rise as a value-commerce alternative is squeezing margins at the lower end.

China was the unambiguous failure. Amazon entered in 2004 by acquiring Joyo.com and exited in 2019^[16] with 0.6% market share — obliterated by Alibaba’s Tmall at 61.5% and JD.com at 24.2%^[17] (TI Insight). The company never adapted to local payment systems, underestimated the importance of seller relationships, and refused to engage in the price wars that define Chinese ecommerce.

Southeast Asia remains marginal. Amazon’s $9 billion regional infrastructure investment announced in 2024 is focused primarily on AWS data centres, not ecommerce. The Singapore marketplace presence hasn’t translated into regional competitiveness against Shopee and Lazada, and it remains to be seen whether Amazon will reboot or leave Southeast Asia for now.

Why can’t Amazon crack South Korea?

South Korea should have been perfect for Amazon. It’s the world’s fifth-largest ecommerce market,^[18] with high smartphone penetration, dense urban populations, and consumers who spend more online per capita than nearly any country on earth. Amazon never seriously tried.

By the time Amazon considered a proper Korean launch, Coupang had already built the logistics moat. Over 100 fulfilment centres, a proprietary last-mile fleet, and Dawn Delivery had set consumer expectations at a level Amazon couldn’t match without billions in upfront infrastructure investment (Korea Herald). South Korea’s 24/7 urban culture — where ordering dinner at 11pm and expecting it before breakfast is normal — required a delivery network purpose-built for the country’s geography and lifestyle. Amazon’s global template doesn’t bend that far.

There’s a cultural dimension too. Korean consumers are intensely brand-loyal within digital platforms and deeply habituated to local payment ecosystems, loyalty programmes, and customer service norms. Naver Shopping and Coupang between them account for over 50% of product discovery, leaving little oxygen for a foreign entrant. A competitor entering today would need to invest billions into logistical and technical infrastructure with no guarantee of cracking the duopoly (Quartr). Amazon, pragmatically, chose not to. Coupang Logistics Model: How South Korea’s Amazon Built a Same-Day Delivery Empire

Could Coupang expand beyond Korea?

This is where the story gets interesting. Coupang’s Developing Offerings segment — which includes Taiwan operations and luxury fashion platform Farfetch — is expected to lose between $950 million and $1 billion in adjusted EBITDA in 2026 (Coupang Q4 2025 Earnings Call). That’s a deliberate bet, not a failure.

Taiwan is the proving ground. Revenue there grew at triple-digit rates through 2025,^[19] with customer growth and retention outpacing the early years of Coupang’s Korean buildout. Coupang’s own last-mile network now covers 70% of Taiwan’s geography^[20] and handles 75% of volume as next-day delivery (Taiwan Business TOPICS). The target is 230 of Taiwan’s 260 cities by 2027. Analysts expect the Taiwan segment to reach EBITDA breakeven by late 2026 or early 2027.

The Farfetch acquisition — $500 million in January 2024^[21] for the distressed luxury marketplace — is the wilder swing. By Q1 2025, management reported Farfetch was nearing breakeven, a significant turnaround for a business that had been haemorrhaging cash. The logic is that Coupang’s logistics expertise and operational discipline can fix what Farfetch’s previous management couldn’t: unit economics.

The question is whether Coupang’s playbook — own the infrastructure, deliver faster than anyone, obsess over the last mile — can translate beyond a small, dense, wealthy country. Taiwan’s early results suggest it can. But Southeast Asia, India, or Japan would be entirely different propositions, each with entrenched incumbents and complex logistics landscapes.

What’s the outlook?

Two paths are diverging. Coupang is betting that going deep beats going wide — that owning 25% of Korean ecommerce, building a second fortress in Taiwan, and running a luxury fashion turnaround is a more defensible business than spreading resources across a dozen countries. The company’s 2026 guidance implies 5-10% revenue growth, disciplined domestic expansion, and continued investment in Developing Offerings.

Amazon is betting the opposite — that its platform model, global brand, and AWS-subsidised economics can grind out market share across Japan, India, and possibly Southeast Asia without needing to replicate Coupang’s infrastructure intensity. Japan is already profitable and growing. India is the long-term prize, with Amazon positioning for what could be a $200 billion ecommerce market by 2030.

Neither company is likely to invade the other’s core territory. Amazon won’t build 100 fulfilment centres in South Korea, and Coupang won’t launch a global third-party marketplace. They’ve each found the model that suits their market — and that’s precisely why comparing them reveals so much about what actually works in Asian ecommerce. The logistics-first model wins where density, speed, and consumer expectations align. The platform model wins where scale, selection, and cross-border infrastructure matter more.

Asia’s big enough for both. The real losers are the companies stuck in the middle — with neither Coupang’s delivery obsession nor Amazon’s platform gravity.