Marketplaces, platforms and the new middlemen

Alibaba, through Taobao and Tmall, pioneered a similar model in China years earlier, monetizing not through direct fees on every transaction but through advertising auctions: merchants bid for visibility because organic discovery is nearly impossible at that scale.

Zalando, Europe's largest online fashion platform, transformed from an inventory-owning retailer into a hybrid: its "Partner Program" lets brands list and fulfill their own stock through Zalando's site, paying a commission, while Zalando keeps the customer relationship, the traffic, and the data.

In all three cases, the platform owns the thing that matters most: the customer's attention. Everyone else pays for access to it.

Why sellers accept the toll

This looks like a bad deal for brands. So why do hundreds of thousands of them sign up?

Reach. Amazon has an enormous built-in audience already searching to buy. Building that traffic independently, through a standalone website, would cost far more in customer acquisition than the marketplace fees.

Trust transfer. Buyers trust Amazon's return policy and delivery promise more than an unknown brand's own site. The platform lends its reputation.

Infrastructure. FBA warehousing and two-day shipping would be prohibitively expensive for most small sellers to replicate. Alibaba's logistics arm Cainiao and Zalando's fulfillment centers offer similar advantages.

This is the classic platform economics trade: sellers give up margin and customer ownership in exchange for demand they could not generate alone. Economists call this a two-sided market, a business connecting two distinct groups (buyers and sellers) where the value to each side depends on how many participants are on the other. For background on platform economics, see the OECD's overview of digital platforms and competition.

The power asymmetry

The relationship is not equal, and that is the point.

Amazon can adjust the "Buy Box" algorithm (the default purchase button that decides which seller wins a sale when multiple offer the same product) with no negotiation. It can launch a private-label product, like Amazon Basics, using sales data from the very sellers it competes against. This practice, sometimes called self-preferencing, drew significant regulatory attention.

In 2023, the U.S. Federal Trade Commission (FTC) sued Amazon, alleging it uses anticompetitive tactics to maintain monopoly power, including punishing sellers who offer lower prices elsewhere. In the European Union, the Digital Markets Act (DMA), in force since 2024, designates large platforms as "gatekeepers" and explicitly restricts self-preferencing, forcing companies like Amazon to treat their own products no better than third-party ones in search rankings. Alibaba faced its own reckoning: Chinese regulator SAMR (State Administration for Market Regulation) fined it roughly $2.8 billion in 2021 for antitrust violations related to forcing merchants into exclusive platform arrangements.

Sellers, meanwhile, have limited leverage individually. Collectively, seller associations and advocacy groups have pushed back, but switching platforms is costly: leaving Amazon often means losing most of your visible demand overnight.

Zalando and the European middle path

Zalando's evolution is worth a closer look because it shows a slightly different power balance.

Unlike Amazon, Zalando started as an inventory retailer buying stock from fashion brands outright. Its shift toward a partner/marketplace model was partly defensive: European fashion brands (adidas, Nike, smaller labels) wanted more control over pricing and customer data than a pure wholesale relationship allowed.

Zalando's Partner Program gives brands more visibility into their own sales data than Amazon typically shares, and Zalando has positioned itself as a "platform for the industry" rather than a competitor extracting private-label advantage. This is a strategic choice: European regulators, and European brand culture, are more sensitive to platform dominance, so Zalando has cultivated a less adversarial reputation than Amazon, even as its commission structure serves a similar function.

Who really captures the value

Mapping value distribution in a marketplace transaction:

| Party | Captures |

|---|---|

| Brand/seller | Product margin (shrinking due to fees) |

| Platform | Referral fee, ad revenue, fulfillment fee, data |

| Logistics arm | Delivery fee (often same company, e.g. FBA, Cainiao) |

| Customer | Convenience, price competition, faster delivery |

| Regulator | Sets boundaries on how much power concentrates upstream |

The critical shift: advertising has become a second toll booth. On Amazon, sponsored product placements are now a massive revenue line, separate from referral fees. Sellers effectively pay twice: once to transact, once to be seen. Amazon's advertising business is estimated to generate tens of billions of dollars annually, rivaling the ad revenues of major traditional media companies (Amazon has disclosed this as a reporting segment since 2021; exact figures are in its 10-KKThe average number of new users each existing user generates through referrals. Above 1.0, growth compounds on itself and becomes exponential.Voir la définition complète → filings).

A simplified worked example: a seller with $100 of product sells through Amazon FBA. Referral fee of 15% ($15), FBA fulfillment roughly $6 to $8 for a small item (estimate, varies by size/weight), advertising spendadvertising spendAny media you pay for: display ads, search ads, social ads, and sponsorships. You buy access to someone else's audience on a per-click, per-impression, or flat-fee basis.Voir la définition complète → often 10 to 15% of sales for competitive categories ($10 to $15). Total platform take: roughly $31 to $38 out of every $100 in sales, before the seller's own cost of goods.

Challengers and counter-moves

Not everyone is playing by Amazon's or Alibaba's rules.

Shopify built its business explicitly as the anti-marketplace: infrastructure that lets brands run their own direct-to-consumer (D2C) stores, keeping customer data and full margin, in exchange for a software subscription and payment fee rather than a percentage-of-sales rent. Its rise reflects brand appetite for an alternative to marketplace dependency.

TikTok Shop is an emerging challenger reshaping discovery: it merges content and commerce, letting creators drive impulse purchases directly inside the app, threatening the search-based discovery model Amazon and Alibaba rely on.

Temu and Shein, backed by Chinese manufacturing networks, compress the supply chain further, often selling directly from factory to consumer with minimal traditional retail intermediation, raising fresh regulatory questions in the EU and US about customs duty thresholds and product safety oversight.

🎬 [VIDEO: "How Amazon's Marketplace Actually Works" - youtube.com/@wsj - a concise explainer on third-party seller economics and Amazon's dual role as retailer and rent collector, useful for visualizing the fee stack described above]

Key Takeaways

• Marketplaces (Amazon, Alibaba, Zalando) monetize by renting access to demand, data, and logistics rather than by buying and reselling goods; this shifts margin from sellers to platforms.
• Sellers accept marketplace fees (commonly 30% plus of revenue when combining referral, fulfillment, and advertising costs, figures are estimates and vary by category) because independent customer acquisition is often costlier.
• Regulators are actively rebalancing power: the EU's Digital Markets Act restricts self-preferencing by "gatekeeper" platforms, the US FTC has sued Amazon over anticompetitive practices, and China's SAMR has already fined Alibaba over exclusivity practices.
• Advertising has become a second, less visible toll on sellers, often exceeding traditional commission fees in competitive categories.
• Challengers like Shopify (control), TikTok Shop (discovery), and Temu/Shein (supply chain compression) show that the marketplace model itself is now under competitive and regulatory pressure, not a permanent endpoint.