Internet marketplaces and network effects

What makes a marketplace a marketplace

A marketplace is a business that connects two or more distinct groups, typically buyers and sellers, and facilitates transactions between them without necessarily owning the inventory or providing the service itself. This is a meaningfully different model from a company that sells its own product or service directly, and interviewers expect you to keep the distinction sharp, because marketplace economics and marketplace risk look nothing like a typical product company's.

The revenue mechanism in a pure marketplace is the take rate: the percentage of each transaction's value that the marketplace keeps as its fee for connecting the two sides and, often, for handling payments, trust and safety, and dispute resolution. A marketplace with a 10% take rate on $100 of transaction volume keeps $10 as revenue. Because the marketplace does not own the underlying inventory, its revenue is a fraction of the total transaction value flowing through the platform, which is exactly why gross merchandise value (GMV), the total dollar value of all transactions completed on the platform, overstates what the marketplace itself actually earns and should never be confused with revenue in a valuation discussion.

GMV versus revenue: the metric interviewers check first

GMV is a useful growth and scale indicator, since it captures the total economic activity a marketplace is facilitating and tends to correlate with how valuable the platform has become to both sides of the market. But GMV is not revenue, and treating it as though it were is one of the fastest ways to signal you have not thought carefully about marketplace economics in an interview. Revenue is GMV multiplied by the take rate, so two marketplaces with identical GMV can have very different revenue and very different valuations if their take rates differ, and a marketplace growing GMV quickly while its take rate is under pressure (from competition, from large sellers negotiating better terms, or from a shift in transaction mix toward lower-take-rate categories) may not be growing revenue nearly as fast as the GMV headline suggests. A careful analyst always asks what take rate a GMV figure implies, and whether that take rate is trending up, flat, or down, before drawing any conclusion about the health of the business.

Network effects: why marketplace leaders are hard to dislodge

Network effects describe a dynamic where a product or platform becomes more valuable to each user as more users join, and they are the central reason a market-leading marketplace or platform tends to be extraordinarily durable once it reaches real scale, which is a large part of why interviewers care so much about identifying and evaluating them.

Direct network effects occur when more users of the same type make the product more valuable to other users of that same type, most classically in communication and social platforms, where a network is more useful to any one person the more of their contacts are also on it. Indirect, or two-sided, network effects are the defining dynamic in a marketplace specifically: more sellers make the platform more attractive to buyers (more selection, more competitive pricing), and more buyers make the platform more attractive to sellers (more potential customers, more liquidity for their listings). This two-sided reinforcement is what creates the classic "chicken-and-egg" problem every marketplace faces at launch, since a marketplace with no sellers cannot attract buyers, and a marketplace with no buyers cannot attract sellers, which is why early-stage marketplace strategy often involves manually or artificially seeding one side of the market before the natural network effect can take over.

Data network effects are a third, related pattern common to internet and advertising businesses specifically: more usage generates more data about user behavior and preferences, which can improve the product (better recommendations, better ad targeting), which attracts more usage, which generates more data. This effect can compound in a way that is somewhat independent of the two-sided marketplace dynamic above, and it is part of why scale in consumer internet businesses tends to be self-reinforcing even outside a strict buyer-seller marketplace structure.

Why network effects matter for valuation, not just strategy

A business with strong, genuine network effects tends to earn a valuation premium relative to a business without them, because network effects function as a moat: a well-funded new entrant cannot simply copy the product and win share, since the incumbent's advantage comes from the size of its existing network, not just the quality of its product in isolation. This is a meaningfully different competitive dynamic than the one that applies to most software companies, where a fast-moving competitor with a genuinely better product can win customers away over time even against an incumbent with more revenue today, discussed in the retention-driven competitive logic covered in the SaaS metrics bankers actually use. An interviewer asking you to evaluate a marketplace or platform business is very often really asking you to assess how real and how durable its network effects actually are, since that assessment matters more to the long-term investment case than almost any near-term financial metric.

It is worth being skeptical here too, because "network effects" is one of the most overused phrases in technology investing, applied loosely to businesses that do not actually have them. A company that simply has many customers does not necessarily have a network effect; the test is whether an additional user genuinely makes the product more valuable to existing users, not merely whether the company has grown large. Being able to articulate why a specific business does or does not have a real network effect, rather than asserting it exists because the company operates online, is exactly the kind of precise thinking that separates a strong answer from a weak one.

Disintermediation risk: the threat unique to marketplaces

A risk specific to marketplaces, without a close equivalent in software or semiconductors, is disintermediation: buyers and sellers who meet through the platform deciding to transact directly with each other outside it, avoiding the take rate entirely on future transactions. This risk is highest when the marketplace's main value is simply making an introduction, and lowest when the marketplace also handles ongoing value that is hard to replicate off-platform, such as payments processing, trust and safety, dispute resolution, logistics, or repeat discovery that keeps bringing new counterparties to each side. A well-run marketplace typically invests deliberately in making itself sticky beyond the introduction, precisely to defend against this risk, and an interviewer evaluating a marketplace business will often ask directly what stops buyers and sellers from cutting the platform out once they have found each other, since the answer says a great deal about how durable the take rate actually is over the life of a relationship rather than just its first transaction.

Advertising-based internet platforms

Not every internet business fits the buyer-seller marketplace model. A large share of the internet sub-sector, referenced in the sub-sector map, monetizes attention rather than transactions: an advertising-supported platform earns revenue by selling advertisers access to its users' attention, typically priced on some combination of impressions delivered and the platform's ability to target the right audience for a given ad. These businesses often exhibit the data network effect described above (more usage generates more behavioral data, which improves targeting, which makes the platform more valuable to advertisers) rather than the classic two-sided marketplace effect between buyers and sellers.

Valuing an advertising-supported platform leans on a related but distinct set of metrics: user growth and engagement (since attention is the underlying resource being monetized), and average revenue generated per user, which captures how effectively the platform converts that attention into advertiser spend. The same discipline that applies to a marketplace's take rate applies here too: user growth alone is not the same as revenue growth, and a platform growing its user base while monetization per user stalls or declines deserves the same scrutiny as a marketplace whose GMV is outpacing its revenue. Building a coherent view on either type of internet business, marketplace or advertising-supported, is exactly what pitching a tech stock in a TMT interview walks through in more depth.

How marketplaces get valued

Because take rate determines how much of GMV the marketplace actually captures as revenue, marketplaces are valued on a multiple of revenue or, in some cases, gross profit rather than GMV, following the same broad logic as how software companies are valued: a marketplace still reinvesting heavily to grow GMV and defend or grow its take rate may show suppressed near-term profitability, in which case a growth-adjusted revenue multiple is the more useful lens, while a mature, more profitable marketplace can be evaluated more conventionally on EV/EBITDA. Take rate trend is one of the first things a banker checks when comparing marketplace peers, since a rising take rate on growing GMV is a much stronger signal than flat or growing GMV accompanied by a shrinking take rate, even if both scenarios could produce similar revenue in a given period.

Practice question

A marketplace grew GMV by a large amount last year, but revenue grew much more slowly. What would you want to know?

The gap between GMV growth and revenue growth tells me the take rate is likely compressing, since revenue is just GMV multiplied by the take rate, so I'd want to understand why. It could be competitive pressure, a rival marketplace undercutting on fees to win share. It could be mix shift, growth concentrated in categories or large sellers that negotiate lower take rates than the platform's historical average. Or it could be a deliberate strategic choice, the company lowering its take rate on purpose to accelerate GMV growth and strengthen its network effects, betting that a larger, more liquid marketplace is worth more long-term even at a lower near-term take rate. Those three explanations lead to very different conclusions about the business. Competitive erosion is a real warning sign about the durability of the marketplace's position. A deliberate growth-over-margin tradeoff can be a completely reasonable strategy, similar in spirit to a software company suppressing EBITDA to fund customer acquisition, as long as the two-sided network effect is genuinely strengthening as a result. I'd look at whether the number of active buyers and sellers is growing, whether liquidity and repeat usage are improving, and how take rate has trended over several periods, not just the last one, before deciding which story actually explains the gap.

What the interviewer is listening for: Whether you immediately reach for take rate as the variable connecting GMV and revenue, and whether you can distinguish a deliberate, strategically sound take-rate tradeoff from a defensive one caused by competitive pressure.