Marketplace vs E-Commerce Store

By Brian Tr · Updated Oct 6, 2026

A marketplace connects third-party sellers with buyers and earns money from the transactions between them. An e-commerce store sells products itself, either from inventory it owns or from a supplier it orders from on demand. Both show a product grid and a checkout button, so they look alike to a shopper. Underneath, they are different businesses with different balance sheets, different risks and different ways of growing.

The short version: a store owns the sale, and a marketplace hosts it. That one distinction decides who holds inventory, who sets the price, how revenue is counted, where margin comes from, and what limits growth. This article walks through each of those differences, shows how a hybrid model blends them, and ends with a rough guide to which model fits which kind of founder.

It's a reference piece, not a playbook. For the definition of a marketplace and the types that exist, start with what a marketplace business is.

The two models in one paragraph each

E-commerce store (first-party, or 1P). The business buys or makes goods, lists them, sets the price, takes the order and ships it (or has a supplier ship it on its behalf). The customer buys from the store. Our overview of e-commerce business models covers the variations, including dropshipping, where the store never touches the stock.

Marketplace (third-party, or 3P). The business builds a platform where independent sellers list their own products or services. Buyers purchase from those sellers through the platform. The marketplace brings demand, trust and payment tools, and takes a share or a fee. The marketplace business model article covers the revenue mechanics in more depth.

Inventory and working capital

This is the biggest structural gap.

A store that holds inventory has to pay suppliers before it gets paid by customers. Stock sitting in a warehouse is cash that can't be spent on anything else, and unsold stock turns into a markdown or a write-off. The more SKUs the store carries, the more cash it ties up. Good inventory management becomes a core competency, not a back-office detail.

A marketplace usually holds none of that. Sellers carry their own stock and their own risk. The platform can grow its catalog without financing it. That's one reason marketplaces can list millions of items with far less capital than a store would need to stock the same range.

The picture isn't one-sided, though. Marketplaces often take on other kinds of working capital risk: payments they collect on behalf of sellers and pay out later, refunds and chargebacks, and any fulfillment service they choose to run. A marketplace that offers its own warehousing and shipping starts to resemble a store on the cost side, even though the goods still belong to someone else.

Dropshipping sits in between. The store takes the order and sets the price, but a supplier holds and ships the stock. Working capital is light, yet the store still owns the customer promise.

Who sets the price and controls the experience

In a store, the business decides the price, the product page, the returns policy, the packaging and the tone of every message. That control makes the experience consistent. It also means every mistake is the store's alone.

In a marketplace, sellers usually set their own prices and write their own listings, within rules the platform sets. The platform controls the layer around them: search, ranking, reviews, checkout, dispute handling and trust signals. This is why quality control is the central operating problem of a marketplace. If one seller ships late or lists a fake product, the buyer blames the platform.

Some marketplaces tighten control to protect the experience. A managed approach, where the platform vets sellers or takes part in fulfillment, trades some of the model's lightness for reliability. We cover that middle ground in managed marketplaces.

Customer ownership

A store owns the customer relationship. It collects the email, sees the full order history and can market directly. The customer thinks of the store's brand.

A marketplace owns a relationship with the buyer too, but the seller has its own claim. Buyers often remember the platform rather than the seller, which is the marketplace's strength. But the platform has to manage the risk of buyers and sellers moving a repeat transaction off the platform to avoid fees. That tension, usually called disintermediation or leakage, has no equivalent in a first-party store.

Revenue recognition: gross versus net

Accounting rules turn the structural difference into a number on the income statement. This is where many founders get confused, because the same sale of a $100 item can show up as $100 of revenue or as a few dollars of revenue, depending on the model.

Under the US revenue standard (ASC 606), the question is whether the company is a principal or an agent for each specified good or service. Deloitte's accounting roadmap puts it this way: a principal recognizes revenue at the gross amount it is entitled to from its customer, while an agent presents revenue at the net amount it retains. The test is control: an entity is a principal if it controls the specified good or service before it's transferred to the customer.

The standard points to indicators of control. Per the same Deloitte roadmap, they include whether the entity is primarily responsible for fulfilling the promise, whether it carries inventory risk before or after the transfer (for example, through a right of return), and whether it has discretion in setting the price. Those indicators aren't a checklist, and the conclusion rests on the facts of each arrangement.

Map that to the two models:

• A store that buys and resells goods typically controls them, carries inventory risk and sets the price. That points to principal treatment and gross revenue.
• A marketplace where sellers own the goods and set prices typically points to agent treatment. The platform recognizes its commission and fees, not the full purchase price.

Amazon shows both treatments inside one company. In its 2025 annual report, it says that for third-party seller services it is not the seller of record in these transactions, and that the commissions and related fulfillment and shipping fees it earns are recognized when the services are rendered. Retail sales, by contrast, are recognized when control of the goods transfers to the customer.

The same question exists internationally. IFRS 15 also deals with principal-versus-agent assessments, and the IFRS Interpretations Committee has published an agenda decision on a software reseller case listed alongside the standard. The details differ in wording, so a company reporting under IFRS should read the standard text rather than assume the US answer carries over.

One caution: a marketplace's accounting conclusion isn't guaranteed to be "agent" just because it's a marketplace. If the platform takes control of goods, sets prices or is the party responsible to the buyer, the facts can change the answer. And a store can be an agent for some items, such as a third-party warranty it merely arranges. The analysis runs per specified good or service.

GMV versus revenue

Marketplaces talk about GMV (gross merchandise volume), and it's easy to mistake it for revenue. They aren't the same thing.

eBay defines GMV in its annual report as the total value of all paid transactions between users on its Marketplace platforms during the period, inclusive of shipping fees and taxes, without adjustment for returns or cancellations. It's a measure of activity on the platform. Net revenue is what the platform keeps.

The gap between them is the take rate. In the same filing, eBay reports 2025 GMV of $79,609 million, net revenues of $11,100 million and a take rate of 13.94 percent, where take rate is net revenue as a share of GMV. Note that eBay's net revenue includes advertising and other income as well as transaction fees, so take rate here is a blended measure.

For a store, none of this applies in the same way. If the store is the principal, its revenue is the full sale price, and the number most people watch is gross margin: what's left after the cost of goods. A marketplace has a much smaller revenue line relative to the volume that moves through it, so comparing "revenue" across the two models can mislead. A marketplace with $1 billion of GMV and a modest take rate and a store with $1 billion of revenue are not businesses of the same size.

Margins and unit economics

The two models earn money in different places.

A store's margin is the spread between what it pays for goods and what it charges, minus fulfillment, returns, payment fees and marketing. Each unit sold carries product cost, so margin percentages tend to be set by category and purchasing power. Scale helps by improving supplier terms.

A marketplace's margin comes from the take rate and any extra services, such as advertising, payments or logistics, applied to each transaction. It has no cost of goods for items sellers own. Its costs are technology, trust and safety, support, payment processing and demand generation. The same fixed platform cost can serve more transactions, which is why marketplaces can show strong operating leverage once they pass a certain volume.

Both models still live or die on unit economics: what each order earns after variable cost, and what it costs to acquire the customer and the seller. A marketplace has to check this on two sides at once, because it pays to acquire both buyers and sellers.

Scaling constraints

Each model hits a different wall.

The marketplace's problem is often called the cold start or chicken-and-egg problem: buyers come only if there's supply, and sellers come only if there's demand. A store doesn't have this problem, because it creates its own supply. A store has the opposite problem: supply is expensive to build, so growth has to be funded.

The hybrid model: 1P plus 3P

Plenty of companies run both models together, selling their own goods and hosting outside sellers on the same site. Amazon is the best-documented case.

Amazon's quarterly earnings release reports a figure called worldwide seller unit mix: third-party sellers' share of paid units, excluding Whole Foods Market. In its release for the second quarter of 2026, dated July 30, 2026, the figure was 61 percent. The same table shows 60 percent for the first quarter of 2026 and 61 to 62 percent across the four quarters of 2025, so the share has been steady rather than climbing.

Revenue tells a different story from units, which is the gross-versus-net point in action. In the 2025 annual report, Amazon's third-party seller services line was $172,162 million out of $716,924 million of consolidated net sales, about 24 percent. Online stores were $269,287 million. So sellers account for a majority of units but a minority of reported revenue, because the third-party line counts commissions and fees rather than the full sale price.

Why would a store add a marketplace? It widens the catalog without buying more stock, brings in demand data about what shoppers want, and adds a fee-based revenue stream. Why would a marketplace add first-party goods? It can fill gaps in supply, guarantee quality on key products and control the experience where reliability matters most. The cost is complexity: the company now competes with its own sellers, runs two sets of operations and has to manage the trust problem that comes with it. Our article on multi-channel marketplace strategy looks at the seller side of that relationship.

A brand selling on marketplaces is not a marketplace

One more distinction trips people up. A brand that lists its products on someone else's marketplace is still an e-commerce business. It's a seller, not a platform. It owns the inventory, sets or negotiates the price and carries the risk, while the marketplace provides traffic and checkout in exchange for fees.

Selling on a marketplace is a channel decision. Building a marketplace is a business model decision. They need different skills: the first is about listings, pricing and fulfillment on someone else's rules, and the second is about designing rules for other people's businesses.

Side-by-side comparison

Which model tends to fit which founder

There's no universally better choice. The fit depends on what the founder has.

• A founder with a product, a brand or a sourcing advantage usually fits a store. Control of the product and the customer is the asset.
• A founder with deep access to one side of a fragmented market (for example, a network of suppliers or an audience of buyers who can't easily find each other) is the usual profile for a marketplace. The scarce thing is the network, not the goods.
• A founder with limited capital may lean toward dropshipping or a marketplace, since neither requires buying stock up front. Each trades capital for a different problem: thin margins and supplier dependence in one case, the cold start in the other.
• A founder who needs tight quality or a consistent experience may prefer a store, or a managed marketplace that adds vetting and fulfillment.
• A founder who already has a store with traffic may consider adding third-party sellers, with the caveats about competing with them.

Treat this as a way to frame the decision, not a rule. Many companies start as one model and add the other later. The article on managed marketplaces goes further on one way the model adapts.

Key Facts: Marketplace vs E-Commerce Store

• Under ASC 606, a principal recognizes revenue at the gross amount, and an agent at the net amount it retains. The test is whether the entity controls the specified good or service before transfer (Deloitte DART).
• Indicators of control include primary responsibility for fulfillment, inventory risk and pricing discretion, and they aren't a checklist (Deloitte DART).
• eBay defines GMV as the total value of all paid transactions between users, including shipping fees and taxes, without adjustment for returns or cancellations (FY2025 10-K).
• eBay's 2025 GMV was $79,609 million, net revenues $11,100 million, and take rate 13.94 percent (FY2025 10-K).
• Amazon says it is not the seller of record for third-party seller services and recognizes the commissions and fees it earns (FY2025 10-K).
• Third-party sellers made up 61 percent of Amazon's worldwide paid units in Q2 2026, excluding Whole Foods Market (earnings release, July 30, 2026).
• Amazon's 2025 third-party seller services revenue was $172,162 million of $716,924 million in consolidated net sales, about 24 percent (FY2025 10-K).

Related Reading

• What Is a Marketplace Business?
• Managed Marketplace
• Marketplace Business Model
• E-Commerce Business Models
• Multi-Channel Marketplace Strategy
• Dropshipping Model
• Unit Economics