Marketplace: from Amazon Marketplace (2000) and Alibaba to the platform economy of 2026, and how to decide whether to sell there

A marketplace is a digital platform where multiple sellers offer products to buyers within a single shopping environment. The platform facilitates discovery, payment, logistics (sometimes), and trust, but it typically does not own the inventory it sells: that role belongs to the third parties who list their products.

The concept sounds simple, and it has become central to commerce on the internet. What is rarely explained in depth is how the model was built, what particular economic dynamics govern it, and what real decision a brand makes when it chooses to sell on a marketplace versus on an owned channel. This guide covers those three aspects with the seriousness they deserve, because the marketplace decision has lasting consequences for margin, brand control, and the customer relationship.

The origin: Amazon, 1995, online bookstore

The modern commercial history of the marketplace begins in July 1995, when Jeff Bezos launched Amazon.com from his garage in Bellevue, Washington, selling books online. Initially Amazon was traditional e-commerce with its own inventory: Amazon bought books, stored them, and sold them. It was not a marketplace.

The fundamental transformation happened five years later. In November 2000, Amazon launched Amazon Marketplace, allowing third parties to sell products—initially used books—through Amazon's platform. This move radically changed the model: Amazon stopped being just a retailer and also became a commerce platform. The company charged a commission on each third-party sale, retained the relationship with the end customer (who saw the transaction as "buying from Amazon"), and expanded the catalog without having to buy inventory.

The decision was initially controversial internally. Allowing third parties to sell alongside Amazon competed with Amazon's own products. But Bezos defended the model, arguing that the customer should have maximum choice; if a third party sold better than Amazon, the customer should be able to buy it. This philosophy became part of Amazon's cultural DNA ("customer obsession").

Amazon Marketplace grew dramatically in the following decades. In 2026, more than 60% of the units sold on Amazon come from third-party sellers, not from Amazon directly, according to the group's official reports. Amazon's third-party arm is, by itself, one of the largest businesses in global e-commerce.

Alibaba, Taobao, and the Chinese model

While Amazon was building its model in the United States, another player was beginning to define marketplace economics in Asia. Alibaba Group was founded in 1999 by Jack Ma and a group of partners in Hangzhou, China. Initially a B2B marketplace connecting Chinese manufacturers with international buyers (Alibaba.com), the company diversified quickly.

In May 2003 it launched Taobao, a consumer-to-consumer and later B2C marketplace that became the Chinese equivalent of eBay and Amazon combined. Taobao introduced innovative payment models (Alipay, escrow to solve the trust problem in online transactions) and social dynamics (direct chat between buyer and seller, behavior-based recommendations).

In 2008 it launched Tmall, in a way separating premium B2C commerce from the open marketplace. By 2026, Alibaba operates the largest marketplace ecosystem in the world measured by GMV (Gross Merchandise Volume), with figures that typically exceed Amazon's. Events such as Singles' Day (every November 11, "11.11") are unique commercial phenomena that move tens of billions of dollars in a single day.

The importance of mentioning Alibaba: the marketplace model has cultural and commercial variants that developed in parallel and have influenced one another. The global platform economy is not just Amazon; understanding Alibaba is necessary to understand the sector.

Other iconic cases: Mercado Libre, eBay, Etsy

Mercado Libre was founded in 1999 in Buenos Aires by Marcos Galperín while he was studying for an MBA at Stanford. Initially similar to eBay (a C2C auction marketplace), it evolved into a full ecosystem: Mercado Pago (payments), Mercado Envíos (logistics), Mercado Crédito (financing), Mercado Ads (advertising). In 2026 it is the largest technology company in Latin America by market capitalization, listed on the Nasdaq since 2007.

eBay, founded in 1995 by Pierre Omidyar, is one of the oldest marketplaces in continuous operation. Originally C2C auctions (covered in C2C), it evolved into a mixed B2C+C2C marketplace with an emphasis on secondhand and specialized products.

Etsy, founded in 2005 in Brooklyn, is a marketplace specialized in handmade and vintage products. Unlike general marketplaces, Etsy has maintained a specific positioning sustained over two decades.

Rakuten in Japan, Shopee in Southeast Asia, Wildberries in Russia, and Trendyol in Turkey are examples of marketplaces that dominate specific regional markets. The geography of e-commerce remains more fragmented than Amazon's global dominance would suggest.

The economic dynamics: take rate, flywheel, dependence

Marketplaces have particular economic properties worth understanding:

Take rate (commission). The percentage the marketplace charges on each transaction. It varies enormously by category, platform, and product type. Typical ranges:

• Amazon: 6-15% depending on category (electronics lower, clothing higher), plus additional fees (FBA storage, FBA fulfillment, etc.) that can raise the total cost to 20-30% of the transaction.
• eBay: typically 10-13% of the final price plus fixed fees.
• Etsy: 6.5% transaction fee + $0.20 listing fee + payment processing.
• Mercado Libre: 10-17% depending on plan and category.
• Booking, Airbnb, Expedia (travel verticals): 15-25% of the reservation.

These take rates are one of the most important levers in the economics of any marketplace seller. A company with a 40% gross margin selling on Amazon with an effective take rate of 20% loses half its margin to the channel. The operational question: does the additional volume make up for the lost margin?

The Bezos flywheel. In 2001, Bezos sketched on a napkin what has become the most famous drawing in e-commerce: the Amazon flywheel. The idea: lower prices attract more customers, more customers attract more third-party sellers, more sellers increase selection, greater selection improves the experience, a better experience attracts more customers—and the wheel spins faster. Additionally, volume enables a lower cost structure, which enables lower prices, feeding the cycle.

This flywheel describes a self-reinforcing competitive advantage that is very hard to overcome once established. It's why so few general marketplaces have been able to compete effectively with Amazon in markets where Amazon already had critical mass.

Multilateral network effects. As covered in platform, marketplaces are two-sided markets where each side is more valuable the larger the other is. Once a marketplace reaches critical mass, attracting competition becomes increasingly difficult.

Power asymmetry. The dominant marketplace has asymmetric power over sellers. It can change terms, modify search algorithms, launch its own products that compete with established sellers, suspend accounts. Sellers can appeal or protest, but typically without real bargaining power.

The real decision: marketplace versus owned channel

For a brand, the decision to sell on a marketplace is not binary—most mid-sized and large brands operate a combination of channels. But the trade-offs are clear and worth spelling out:

What you gain by selling on a marketplace

Instant access to existing demand. The marketplace already has millions of visitors with purchase intent. Launching a product on Amazon can generate first sales within hours; launching an owned channel from scratch can take months to reach comparable volume.

Trust by association. Buyers trust Amazon, Etsy, and Mercado Libre as platforms. That trust transfers partly to the seller, especially for small or new brands without a reputation of their own.

Complete infrastructure. Payment, fraud management, partial customer service, fulfillment (FBA on Amazon), ratings, search system. Replicating all of this on an owned channel costs significantly.

Learning without large risk. Testing a new product on a marketplace has a low opportunity cost compared with building an entirely new channel. Validating demand before investing in an owned channel can be a smart use of the marketplace.

Internationalization capability. International marketplaces (Amazon, eBay, Alibaba) make it easier to enter foreign markets without building a local presence from scratch.

What you lose

Margin. The take rate plus additional fees can consume between 15% and 30% of the sale price, depending on platform and category.

Direct relationship with the customer. The marketplace typically retains the relationship. The brand doesn't get the customer's email, can't do direct email marketing, and can't easily build a recurring relationship.

Customer data. The marketplace sees behavior and purchase patterns that the brand doesn't. That information asymmetry structurally favors the platform.

Experience control. You don't decide how your product page looks, which cross-sells appear next to your product, or where your product ranks in searches. The marketplace decides.

Diluted brand differentiation. Your brand appears next to competitors on the listing page. The differentiation you build on an owned channel is flattened on a marketplace where everyone has a similar format.

Risk of competition from the marketplace itself. Amazon has been accused in multiple antitrust cases of using third-party seller data to inform the launch of its own products (the AmazonBasics product line). Sellers can't control this.

Dependence. If the marketplace changes its algorithm, raises commissions, or suspends your account, the business can collapse overnight. That asymmetric dependence has repeatedly proven risky.

When the answer is a marketplace, when it's an owned channel

The practical reality for brands:

A marketplace works especially well when:

• You sell a product that already has established demand in a known category.
• Your brand is unknown and you need quick validation.
• You operate in a product category where a huge catalog is an advantage (books, consumer electronics, commodity products).
• You have limited logistics capacity and the marketplace's infrastructure covers the gap.
• You want to expand into new geographies without a large investment in local presence.

An owned channel works better when:

• Your brand has a differentiated positioning that is diluted on a marketplace.
• You offer a differentiated customer experience that requires control over the journey.
• Your product has a high LTV and you need a recurring relationship with the customer.
• The extra margin from not paying a take rate justifies the investment in building a channel.
• Your product is premium and association with a generic platform damages perception.
• You want customer data to build a long-term business.

Hybrid model (the most common):

• An owned channel as the main store with a complete brand experience.
• A marketplace as an acquisition and discovery channel, possibly with a limited catalog.
• Cross-promotion: the marketplace generates awareness, the owned channel retains the relationship.

The European and Spanish reality in 2026

It's worth naming some particularities of the context:

Amazon is dominant but not a monopoly. In Spain, Amazon processes a significant but not overwhelming share of e-commerce. The share varies by category—high in electronics and books, lower in fashion where others dominate (Zalando, About You, AliExpress) and very low in certain specialized verticals.

The EU's Digital Markets Act—approved in 2022, in force since 2024 (covered in platform)—has imposed specific obligations on large marketplaces designated as "gatekeepers" (Amazon, Apple, Google, Meta, Microsoft, ByteDance, Booking). These obligations include: no unfair prioritization of their own products, transparency about ranking algorithms, and the ability for users to choose alternatives. Enforcement is active and contested, with significant fines in some cases.

Vertical marketplaces with traction. Wallapop, Vinted (covered in C2C), Hannun (a Spanish sustainable-furniture marketplace), and Etsy in its category. For many brands, specialized marketplaces can be more profitable than general ones.

Emerging market: TikTok Shop, launched in Spain and other European markets in 2024. It combines content and commerce in a distinctive format, still in development. For D2C brands with a strong social audience, it can be a relevant channel.

Affordable logistics. Carriers such as Correos, MRW, GLS, SEUR, and DHL have significantly improved e-commerce logistics in Spain, which has reduced the differential advantage of Amazon FBA. Sellers can achieve 24-48h delivery with their own carriers.

Common mistakes in marketplace strategy

Treating the marketplace as a low-effort channel. "We put products on Amazon" is not a strategy. Optimized listings, review management, stock management, customer service, and internal promotions require ongoing active work.

Not optimizing listings. Title, images, description, attributes, internal keywords—each element affects the marketplace's internal ranking and the conversion rate. Amateur listings compete poorly.

Underestimating the complexity of logistics. If you use FBA, there are strict rules (package preparation, labeling, shipping to Amazon warehouses) that require a process. If you use FBM (Fulfilled by Merchant), you manage shipments directly with all their complications.

Not measuring real profitability by channel. The right question isn't "do they sell on Amazon?" but "how much real margin does Amazon produce after all costs (take rate, FBA, returns, internal advertising) versus an owned channel?". Many brands earn less on a marketplace than they think.

Ignoring Amazon advertising. Being on Amazon without investing in Amazon Ads (Sponsored Products, Sponsored Brands, etc.) means low organic visibility in competitive categories. Amazon's economics push toward paid advertising.

Over-dependence on a single marketplace. If Amazon is 80% of your revenue and Amazon decides something that hurts you, your business suffers. Diversifying channels is protection.

Not protecting the brand on a marketplace. Counterfeits, copycats, and unauthorized sellers of your brand are real problems on large marketplaces. Programs like Amazon Brand Registry or IP protection on other platforms are a necessary investment.

Confusing take rate with total cost. Beyond the take rate, there are fees for listing, fulfillment, advertising, returns, and inventory storage. The total cost can be significantly higher than the nominal take rate.

Not preserving the customer relationship when it is possible. Although a marketplace typically retains the relationship, there are legitimate tactics: inserts in the packaging that invite customers to sign up for email, post-sale programs that bring the customer to the owned channel for complementary services.

Marketplace and creative operations

For a brand that sells on a marketplace, the production of visual and text assets is critical. Each listing needs professional product images, a description optimized for keywords and conversion, a product video where applicable, infographics to show attributes, A+ content (Amazon) or equivalents to differentiate. Without that production, listings compete at a disadvantage against sellers who do invest.

That coordinated production connects with creative operations: content production generates assets for multiple marketplaces, brand management ensures visual coherence despite each platform's restrictive formats, and creative KPIs measure the impact of listing improvements on conversion rate.

At Polimake, that logic lives in three surfaces: Studio coordinates multi-marketplace campaigns, Studio produces variants optimized for each platform, Media stores product images, demo videos, and reusable listing templates.

---

If you lead product, marketing, or e-commerce and you've arrived here looking for an answer about marketplaces, the most useful thing you can take from this article is probably the combination of three ideas: a marketplace and an owned channel are not alternatives, they are complementary channels (most serious brands operate both), the effective take rate is typically higher than the nominal one (when all fees, advertising, and operating costs are included), and dependence on a marketplace is a structural risk that deserves conscious management (diversify channels, maintain the customer relationship when possible, monitor policy and algorithm changes).

To round it out, C2C covers the model where the platform facilitates transactions between individuals, D2C covers the opposite model of a pure owned channel, and platform covers the underlying concept that governs multi-sided dynamics.

Quick references

• C2C (Consumer to Consumer) — the parallel model where the platform facilitates between individuals.
• D2C (Direct to Consumer) — the opposite model of a pure owned channel.
• Platform — the underlying concept.
• Competitive analysis — to evaluate the ecosystem of marketplaces relevant to your sector.
• Conversion funnel — how the audience moves between channels.