Unit economics for a marketplace: the three-sided contribution margin

Key Takeaways

• Take rate, the percentage of transaction value the marketplace retains, is the metric that determines whether growth in GMV (gross merchandise value) actually benefits the platform, GMV growth alone says nothing about profitability.
• Contribution margin equals take-rate revenue minus variable cost per transaction; a marketplace with high GMV and negative contribution margin per transaction is scaling its losses, not its business.
• CAC must be split by side, buyer CAC and seller/supply CAC behave completely differently and must be tracked independently, since blending them hides which side is actually expensive to acquire.
• A 3:1 LTV:CAC ratio is commonly cited as a viability threshold for marketplaces too, but for a genuinely multi-sided platform (buyer, seller, and a third participant, such as a service provider or logistics partner) that single ratio has to be checked on every side separately, not once for the platform as a whole.

A marketplace headline metric like "GMV up 40% year over year" answers a growth question, not a profitability question, and the two can point in completely different directions if the transaction economics underneath that GMV are negative. For a genuinely three-sided marketplace, buyers, sellers, and a third participant such as a logistics or service partner, the unit economics question isn't one calculation, it's at least three, run independently.

Take rate: the number GMV growth doesn't tell you

Take rate is the percentage of transaction value the marketplace retains, and it's the critical unit-level metric determining whether GMV growth actually translates into revenue the platform benefits from (TMT IB Guide, marketplace economics: GMV, take rate, and unit economics, retrieved 2026-09-08). GMV is the total value flowing through the platform; take rate determines what fraction of that value the marketplace itself actually captures as revenue. A marketplace can grow GMV substantially while its own revenue barely moves, if take rate is thin or being discounted to win volume, which is why GMV alone, without take rate alongside it, is close to meaningless as a health indicator.

Contribution margin: where the actual profitability question lives

Contribution margin equals take-rate revenue minus the variable cost incurred per transaction, payment processing, support, fraud/dispute handling, any logistics cost the platform absorbs (TMT IB Guide, retrieved 2026-09-08). A marketplace with high GMV and negative contribution margin per transaction isn't scaling a business, it's scaling losses at an increasing rate, since every additional transaction adds to the loss rather than to profit (TMT IB Guide, retrieved 2026-09-08). This is the single most important check before treating GMV growth as good news: is contribution margin per transaction positive, and is it improving or eroding as volume grows. A platform subsidising transactions to win share can look identical to a genuinely healthy one on a GMV chart while being on completely opposite trajectories underneath.

Why CAC has to be split by side, not blended

Unit economics on one side of a marketplace directly affect unit economics on the other, a cross-side relationship specific to multi-sided platforms, more supply attracts more demand and vice versa (Vibrating Melon, unit economics of two-sided marketplaces, retrieved 2026-09-08). CAC must be split by side: buyer CAC and seller CAC behave differently and must be tracked independently, since blending them into a single "marketplace CAC" hides the true cost of acquiring supply versus demand (Vibrating Melon, unit economics of two-sided marketplaces, retrieved 2026-09-08).

For a three-sided model, this splits three ways, not two. If a platform connects buyers, sellers, and a fulfilment or verification partner, each side has its own acquisition cost, its own retention curve, and its own contribution to the transaction's overall economics. A common failure mode is subsidising one side heavily to bootstrap volume (a standard, often necessary marketplace strategy in the early stage) without tracking that side's CAC and contribution margin explicitly, which makes it impossible to know when the subsidy needs to end or how much runway it's actually consuming. Run each side's CAC and contribution figures through the CAC/LTV calculator separately rather than blending them into one platform-wide number.

Applying the 3:1 threshold to a multi-sided platform

A 3:1 LTV:CAC ratio is commonly cited as a viability threshold for marketplaces, the same way it is for SaaS, and it remains one of the most important single ratios in marketplace unit economics (TMT IB Guide, retrieved 2026-09-08). For a genuinely three-sided platform, though, that ratio needs checking on each side independently rather than once for the blended platform. A platform can show an overall 3:1 ratio while one side, say, supply acquisition, sits well below 1:1, subsidised entirely by an overperforming demand side. That's not automatically wrong (early-stage supply subsidisation is a standard playbook), but it needs to be a visible, deliberate choice rather than a fact hidden inside a blended average that looks healthy on paper.

Building the model correctly from the start

For an idea still being validated, the practical sequence is: estimate take rate and per-transaction variable cost first (contribution margin per transaction), then estimate CAC separately for each side the platform depends on, then check whether each side's economics, not just the blended average, clears a sensible LTV:CAC threshold given the platform's stage. This sits squarely inside idea validation work for a marketplace concept, since a business model that only works with negative contribution margin subsidised indefinitely by outside capital is a different, much riskier proposition than one that's genuinely unit-economically sound once past an initial bootstrap phase.

Frequently asked questions

Is negative contribution margin always a red flag for a marketplace?

Not always, early-stage marketplaces often deliberately subsidise one side to bootstrap the network effect. The red flag is negative contribution margin that isn't tracked explicitly, isn't improving over time, or isn't tied to a clear plan for when the subsidy ends.

How do I calculate CAC separately for three different sides?

Track acquisition spend and new participant count independently for each side (buyers, sellers, third-party participants), the same way you would for two entirely separate businesses, then calculate each side's CAC, contribution margin, and LTV:CAC ratio on its own before looking at any blended figure.

Should take rate be the same across all transaction types on the platform?

Not necessarily, take rate often varies by category, transaction size, or which sides are involved in a given transaction. What matters is knowing the actual blended take rate your specific transaction mix produces, rather than assuming a single headline take rate applies uniformly.

The bottom line

GMV and take rate together tell you how much revenue a marketplace is generating; contribution margin, calculated and tracked separately for every side of the platform, tells you whether that revenue is actually profitable at the unit level. A three-sided marketplace that only tracks blended, platform-wide numbers can be scaling a loss on one side while a healthy other side masks it, and the only way to catch that before it becomes a capital problem is to run the unit economics on each side independently.

Figures and definitions were verified on 8 September 2026 against published marketplace unit economics research. Take rate, contribution margin, and CAC benchmarks vary enormously by marketplace category and stage; build your model from your own per-side transaction and acquisition data rather than a generic benchmark.