Skip to main content

The marketplace business model

A marketplace is a platform that connects two (or more) types of participants—buyers and sellers—and takes a fee or commission on transactions between them. The power of the model lies in network effects: the more sellers on the platform, the more valuable it is to buyers; the more buyers, the more valuable it is to sellers. This creates a virtuous cycle where growth accelerates as the network expands. Examples include eBay (auctions), Airbnb (short-term lodging), Stripe (payments), Uber (ride-sharing), Doordash (food delivery), and Amazon Marketplace (retail). Marketplaces can create enormous value because they require no inventory (the goods or services come from sellers), no employees delivering the service (sellers do that), and minimal capital expenditure. But they are also challenging to build because the network effects require simultaneous growth on both the buyer and seller sides.

Quick definition: A marketplace is a platform that connects multiple parties (typically buyers and sellers) and extracts value by taking a fee on transactions or providing infrastructure services. The business model is capital-light but requires achieving scale on both sides of the network to become valuable.

Key takeaways​

  • Marketplaces have extremely high gross margins (often 70–90%) because the platform does not own inventory or directly provide services.
  • The critical metric for a marketplace is liquidity—the ratio of active buyers to active sellers, and the speed at which transactions occur.
  • Network effects are central to marketplace value; a thin, fragmented marketplace is worth little; a thick, concentrated marketplace (where most transactions in a category happen) can be worth billions.
  • Take rate (the percentage of transaction value the platform captures) is a key driver of profitability; a marketplace that cannot raise take rates without losing volume is trapped.
  • The winner-take-most dynamics of marketplaces mean that second or third place has little value; the competitive dynamics are winner-take-most, not winner-take-all.

Why marketplaces are valuable: The power of network effects and capital efficiency​

Marketplaces create value by solving a coordination problem. Before Uber, if you wanted a ride, you either owned a car or called a taxi company. There was friction—you did not know prices, availability, or driver quality. Uber eliminated that friction by creating a platform where drivers and riders could find each other. The value to the rider is convenience (get a ride with a tap); the value to the driver is income (earn money whenever desired). Uber extracts value by taking a commission on each ride.

This model has three characteristics that make it extremely valuable:

1. Capital efficiency: Unlike a traditional taxi company or car rental business, Uber owns no cars and employs no drivers. It is a capital-light business. The company that delivers the service (the driver) bears the capital cost, and Uber captures a percentage of the value created. This allows Uber to scale to billions of dollars in transactions with a tiny balance sheet.

2. Immediate network effects: The value of the network increases with size. A marketplace with one seller is worthless to buyers; a marketplace with one million sellers is extremely valuable. As Uber added more drivers, wait times fell and prices fell (for riders), which attracted more riders, which attracted more drivers. This virtuous cycle is powerful because it is automatic—the platform does not have to do anything to create it; it emerges from the network structure.

3. High gross margins: Because the marketplace does not bear the cost of the service (the seller does), the gross margin is just the transaction fee. If Uber takes 25% of the fare as commission, the gross margin is 25%. Compare this to a taxi company, which bears the cost of the car, fuel, insurance, and driver—gross margins might be 30–40%. Uber's model allows for lower margins and still be profitable on an operating basis much faster.

These three characteristics combine to create value. A marketplace can grow revenue without proportional cost increases; it can achieve high margins while still being competitive; and it can scale with minimal capital. Compare this to a product company (which must manufacture and distribute) or a service company (which must employ people), and the marketplace model is much more efficient.

Two-sided networks: The challenge of bootstrapping​

The challenge of building a marketplace is the chicken-and-egg problem: To attract sellers, you need buyers; to attract buyers, you need sellers. Early-stage marketplaces often have to subsidize one side or both sides to break the deadlock.

Subsidizing demand side (buyers): Offer incentives (discounts, credits, referral bonuses) to attract buyers. Uber did this aggressively, offering $20 credits to new riders. Doordash offered discounts on first orders. This burns cash but creates critical mass on the demand side, which attracts supply (sellers) without additional incentive.

Subsidizing supply side (sellers): Offer incentives (sign-up bonuses, guaranteed earnings floors) to attract sellers. Uber did this with bonuses for drivers hitting certain trip targets. Airbnb did this with professional photography services for hosts. This is expensive but necessary if supply is scarce.

Targeting verticals with existing imbalance: Some marketplaces succeed by targeting a vertical where supply or demand is already concentrated. For example, StockX (sneaker marketplace) targeted sneaker collectors and resellers who were already online and organized; it did not have to recruit sellers from scratch.

Capital-subsidized growth: The venture model of many marketplaces involves massive capital raise upfront, followed by years of subsidizing both sides until critical mass is achieved. DoorDash, Airbnb, and Uber all spent years burning cash to build networks before eventually achieving profitability. Investors bet that the network would eventually create sufficient value to justify the losses. For Airbnb and Uber (post-profitability), that bet paid off. For some other marketplaces (e.g., Groupon), the subsidies were unsustainable and profitability came late, with lower margins than expected.

Liquidity and thickness: The measure of marketplace value​

The ultimate measure of a marketplace's value is liquidity: how quickly and easily can a buyer find and transact with a seller, and vice versa?

A liquid marketplace has:

  • High ratio of sellers to buyers (or vice versa): For each buyer, there are many options; for each seller, there are many potential customers. This creates choice and competition, which is good for pricing and quality.
  • Fast transaction velocity: The time between when a buyer arrives and when a transaction is completed is short. For Uber, this is minutes; for Airbnb, it is days; for eBay, it is hours or days depending on the auction. Faster velocity means the marketplace is handling more transactions per unit of participants.
  • Low friction: The process of matching, negotiating, and completing a transaction requires minimal steps. Uber minimizes friction by one-click booking; Amazon minimizes friction with 1-Click checkout.
  • High repeat rate: Buyers and sellers transact repeatedly, not just once. A buyer that only purchases once generates limited lifetime value; a buyer that purchases weekly (like a DoorDash customer) is highly valuable.

Thickness (a related concept) measures how concentrated transactions are. A thick marketplace for used cars in a city means most used car transactions happen in one place; a thin marketplace means transactions are fragmented across many platforms. Thickness is determined by the size of the buyer and seller bases and the network effects that concentrate them.

The most valuable marketplaces are both liquid and thick. Thick and illiquid (e.g., a marketplace where sellers are abundant but buyers are scarce, or vice versa) has low value. Liquid and thin (e.g., a marketplace with adequate balance on both sides but low overall size) also has low value.

Take rate: The lever of marketplace profitability​

The take rate is the percentage of transaction value that the marketplace captures as commission. It is the primary lever of profitability.

Take rates vary widely by marketplace:

  • High take rate (25–50%): Niche marketplaces, marketplaces with strong network effects and limited competition, or marketplaces providing valuable services. Examples: Uber (25–30%), DoorDash (25–30%), Airbnb (3–16% depending on booking size, plus host service fees), specialized marketplaces.
  • Medium take rate (5–15%): Mature, competitive marketplaces where the marginal cost of the transaction is low but take rate is limited by competition. Examples: Amazon Marketplace (15%), Etsy (6.5%), eBay (12%).
  • Low take rate (<5%): Highly competitive, commoditized marketplaces or marketplaces with thin margins. Examples: payment processors like Stripe (2.9% + $0.30), StockX (sneaker resale, takes a cut on both sides).

A marketplace's take rate is determined by:

  1. Competitive landscape: If there are many competing marketplaces, take rate is limited because sellers and buyers can easily switch. A monopoly marketplace (Amazon Marketplace, eBay in certain categories) can support higher take rates.
  2. Value provided: If the marketplace provides clear value (payments, trust, logistics, customer base), it can support higher take rates. A commodity marketplace (where the marketplace adds no value beyond matching) must have low take rates.
  3. Network depth: A deep network with strong network effects can support higher take rates because switching costs are high. A thin network cannot.
  4. Elasticity of supply and demand: If sellers are price-sensitive and will leave if take rates increase, take rate is capped. If sellers have few alternatives, take rate can be high.

The best marketplaces are those that can raise take rates over time as they build network effects and become more valuable. Stripe raised its take rate from 2.9% + $0.30 to include additional services and products (fraud detection, subscriptions, payouts), increasing the take rate per transaction. Uber raised prices through surge pricing, increasing its take rate. Airbnb increased its take rate by moving from just host commission to also charging guests a service fee.

The worst marketplaces are commoditized, with low take rates, limited ability to increase take rates without losing share, and high competition. An investor should be wary of a marketplace with a take rate under 5%, limited pricing power, and a deteriorating competitive position.

The competitive dynamics: Winner-take-most, not winner-take-all​

A common misconception is that marketplaces are "winner-take-all," where the largest platform dominates and others fail. This is sometimes true but often not. More commonly, they are "winner-take-most," where the largest platform has a dominant share but second or third place can be profitable.

This is because marketplaces are not truly monopolistic; there is supply and demand for multiple platforms. In ride-sharing, Uber dominates in the U.S., but Lyft remains profitable and has a 30% market share. In food delivery, DoorDash dominates, but Uber Eats, Grubhub, and local players remain viable. In e-commerce, Amazon dominates, but Shopify, Etsy, and niche platforms thrive.

Where winner-take-all more closely applies:

  • Payments: The vast majority of online payments go through a few processors (Stripe, Square, PayPal). This is because payment processors are infrastructure; merchants want one processor, not five. Stripe's dominance is enforced by integration friction.
  • B2B enterprise software: Once a company adopts a CRM (Salesforce) or ERP (SAP, Oracle), switching costs are high. New entrants face an uphill battle. Salesforce's dominance in CRM is nearly monopolistic.
  • Search: Google dominates search because of its quality and network effects; there is little room for second place.

Where winner-take-most more closely applies:

  • Ride-sharing: Uber is larger, but Lyft, Didi, and local competitors remain viable because riders and drivers will use multiple platforms, and regulation allows multiple operators in most cities.
  • Food delivery: DoorDash is largest, but Uber Eats, Grubhub, and local players are viable because restaurants and customers use multiple platforms.
  • Short-term rentals: Airbnb dominates, but Vrbo, Booking, and local platforms remain viable because they have different host and guest bases.

The difference often comes down to whether the network effects are truly global and monopolistic (search, payments) or whether they are partial and allow for multiple winners (ride-sharing, delivery, short-term rentals).

Profitability timing in marketplaces: The long road​

Unlike subscription businesses, which can reach profitability fairly quickly (3–5 years) if unit economics are good, marketplaces often take much longer because the path to profitability requires scale. A marketplace with $1 billion in GMV (gross merchandise volume, the total value of transactions) and a 20% take rate has $200 million in gross revenue. But to achieve $1 billion in GMV, the company likely spent 5–10 years and billions of dollars subsidizing both sides of the network. By the time the marketplace is profitable, substantial capital has been deployed.

Airbnb is illustrative. It was founded in 2008, went through years of growth and network building, and went public in 2020 at a $100 billion valuation. It did not achieve profitability until 2021, 13 years after founding. But once profitable, it generated $6 billion in net cash from operations in 2022—the delayed profitability eventually produced very strong cash flows.

This dynamic means that marketplace investors must be patient and capital-aware. An early-stage marketplace that is unprofitable and burning cash is not necessarily a bad investment; it might be building valuable network effects that will eventually produce exceptional returns. But the investor must have conviction that:

  1. The network effects are real (not just theoretical).
  2. The two-sided network can be balanced (buyers and sellers can be recruited without unlimited subsidy).
  3. The take rate can eventually be raised without collapsing the network.
  4. The path to profitability is clear (e.g., reduce subsidies over time, improve take rate, increase transaction velocity).

If any of these conditions fail, the marketplace will never be profitable, and it is a value trap.

The threat to marketplaces: Direct relationships​

A persistent threat to marketplace models is that once a buyer and seller meet on the marketplace, they may transact directly, cutting out the marketplace.

For example, on Airbnb, a guest and host might meet and then communicate directly and book outside the platform to avoid the service fees. On Uber, a rider and driver might exchange numbers and arrange rides directly, cutting Uber out. On Etsy, a buyer might contact the seller directly for future purchases.

Marketplaces defend against this through:

  • Contractual restrictions: Terms of service that prohibit direct transactions or require the platform to take a cut. This is enforceable for businesses but harder to enforce for individuals.
  • Payment processing: Requiring all transactions to go through the platform's payment system, making direct transactions difficult (no invoice, no integrated payment).
  • Reputation and insurance: The marketplace provides buyer and seller protection (dispute resolution, insurance, ratings) that adds value beyond matching. This makes direct transactions less attractive.
  • Lock-in through data: As the platform collects data on buyers, sellers, and transactions, this data becomes valuable and switching to a direct relationship becomes more costly.

The strongest marketplaces have high switching costs through integrations, data lock-in, or structural features that make direct relationships difficult.

FAQ​

How is marketplace value different from product company value? A product company's value is based on its product, brand, customer base, and ability to innovate. A marketplace's value is based on the strength of the network (thickness, liquidity, network effects) and the take rate. A marketplace with a weak network but high take rate is worth less than a marketplace with a strong network and low take rate, because the network can sustain higher take rates over time. A product company's value comes from intellectual property and execution; a marketplace's value comes from the network itself.

What is the difference between take rate and commission? They are often used interchangeably. Take rate is typically the percentage of transaction value; commission is similar. Some marketplaces also charge transaction fees (a flat fee per transaction or a fixed amount plus percentage), and some charge subscription fees to sellers for listing or access. The total "take" is the sum of all these, but take rate usually refers to the main transaction-based percentage.

Can a marketplace be profitable but still destroy value? Yes. A marketplace that is profitable but is losing market share, seeing rising churn on the buyer or seller side, or facing commoditization and falling take rates is destroying value even if it is currently profitable. The opposite is also true: an unprofitable marketplace that is building a strong network, seeing improving retention, and raising take rates may be creating value despite current losses. This is why understanding the trajectory of the network is more important than current profitability.

What is the role of regulation in marketplace profitability? Regulation can significantly impact marketplace profitability. For example, Uber faces ongoing regulatory challenges in many cities regarding driver classification, insurance, and licensing. Higher regulation increases costs and can reduce the network's size. Some marketplaces have benefited from favorable regulation (e.g., Airbnb in cities that do not impose onerous restrictions). An investor should assess regulatory risk when evaluating marketplaces.

How do network effects in a marketplace differ from network effects in other businesses? Network effects in a product (e.g., social media) are direct—the value of the product to each user increases with the number of other users (everyone you want to talk to is on the platform). In a marketplace, network effects are indirect and two-sided—the value to buyers increases with more sellers, and vice versa. This creates unique dynamics; the marketplace is only valuable if both sides are served, not just one. This is why marketplace growth is often slower than pure network products in early stages (you must grow both sides), but can be much more valuable at scale (the network serves a fundamental economic function).

What happens to a marketplace's valuation if take rate falls? It can fall dramatically. If take rate falls because of competition, regulation, or inability to raise prices without losing share, profitability and cash flow both decline. For example, if a marketplace's take rate falls from 20% to 15%, and the marketplace is already profitable, profitability could decline by 25%. The stock would typically rerate downward based on the lower expected free cash flow. This is a key risk for mature, commoditized marketplaces.

Summary​

Marketplaces are capital-light, high-margin business models that create value by connecting buyers and sellers. The critical metrics for evaluating a marketplace are liquidity (fast matching, low friction), thickness (concentrated transaction volume), take rate (profitability per transaction), and network effects (defensibility). The most valuable marketplaces have strong network effects, high take rates, and clear defensibility; the least valuable are commoditized with thin networks and limited pricing power. Profitability timing in marketplaces can be long (7–10+ years) because achieving network scale requires substantial capital deployment on both sides. But once profitable, marketplaces can generate exceptional cash flows due to their capital-light nature. The investor who can distinguish between a marketplace that is building real network effects and one that is wasting capital in subsidies has a significant edge.

Next​

Read The SaaS business model to understand the characteristics of software-as-a-service, a variant of the subscription model with distinct mechanics and economics.