Marketplace Metrics, Explained: GMV, Take Rate, CAC and What They Actually Tell You

There's a particular kind of conversation that happens when two marketplace founders compare notes. One says they're doing $2 million a year. The other nods, impressed. Neither of them has established whether that's $2 million flowing through the platform or $2 million the business actually gets to keep — and those are wildly different companies.

Marketplace metrics get muddled more than most, partly because the vocabulary was borrowed from ecommerce and SaaS, where it means something slightly different. So here's a plain-English pass through the numbers that matter, what each one is actually measuring, and where each one will quietly mislead you.

GMV: the number everyone leads with

Gross Merchandise Value is the total value of transactions completed on your marketplace over a period. If a thousand bookings go through at an average of $120, your GMV is $120,000 for the month.

GMV is a measure of activity, not of business health. It tells you how much commerce your platform is facilitating, which genuinely matters — it's the raw material everything else is calculated from. But it is not money you have. Almost all of it belongs to your sellers.

Two things to watch. First, don't double-count: if you charge both a buyer fee and a seller fee, GMV is still just the transaction value, counted once. Second, decide early whether you're reporting gross or net of cancellations and refunds, and then stay consistent. A GMV figure that quietly includes cancelled bookings will flatter you for exactly as long as it takes someone to check.

Take rate: your slice

Take rate is the percentage of GMV you keep as fees. Charge 12% on that $120,000 and your take rate is 12%.

Most marketplaces land somewhere between 10% and 20%, though it varies enormously by category — high-touch services can support 25% or more, while high-value goods often can't sustain 5%. The number itself matters less than whether it's stable. A take rate that's drifting down usually means your largest sellers are negotiating, or that transactions are getting bigger without your fee structure keeping pace.

Marketplace revenue: what you actually earn

Marketplace revenue — sometimes called net revenue — is GMV multiplied by your take rate. In our example, 12% of $120,000 is $14,400.

This is the line that matters when someone asks how big your business is. The $120,000 is your sellers' revenue. The $14,400 is yours. It's also, incidentally, the number your accountant is likely to put on the top line, because under most reporting standards a marketplace is an agent rather than a principal and reports only the commission.

The gap between the two is why GMV is such a seductive metric. It's the largest number available and it grows fastest. It just isn't the one that pays your team.

Contribution margin: what survives the costs of the transaction

Here's where most founders stop measuring, and it's the most important step.

Contribution margin is what's left of your marketplace revenue after the costs that scale directly with each transaction. Payment processing is the big one. At roughly 2.9% plus 30 cents, a $120 booking costs $3.78 to process — and if you're absorbing that rather than passing it on, it comes out of your $14.40, not out of the $120.

Run that across a thousand bookings and $3,780 of your $14,400 is gone. Add refunds, chargebacks, fraud and the support time a transaction generates, and call it another $1,000. You're left with roughly $9,620.

So: $120,000 of GMV, $14,400 of revenue, about $9,600 of contribution. Eight percent of the headline number. That $9.62 per booking is the money available to pay for everything else — engineering, marketing, and eventually profit.

Every acquisition decision you make should be measured against that figure, not against GMV and not against revenue.

CAC: what a customer costs you

Customer acquisition cost is total spend on acquiring customers divided by the number of customers acquired. Simple in principle, and almost universally calculated wrong.

Three mistakes account for most of it:

• Counting signups instead of customers. Someone who created an account and never booked is not a customer. Divide by people who actually transacted, or your CAC will look excellent right up until you wonder why revenue isn't following.
• Only counting ad spend. Agency fees, creative production, tooling and the salary of whoever runs the channel all belong in there. Fully loaded CAC is often 30–50% higher than the ad platform reports.
• Blending everything together. If organic and referral are cheap and paid is expensive, a blended CAC hides the fact that your paid channel doesn't work. Calculate it per channel as well as blended.

And in a two-sided marketplace, you have two of these. Acquiring a seller costs something quite different from acquiring a buyer, and the two rarely converge. Track them separately.

Say you spend $12,000 in a month and 300 new customers complete a first booking. Your CAC is $40. Against $9.62 of contribution per booking, that customer needs to book roughly four times before you've broken even on acquiring them.

LTV and the ratio that ties it together

Lifetime value is the total contribution margin a customer generates before they stop using your marketplace. Note contribution margin, not GMV — an LTV built on GMV is a fantasy number that will justify almost any spend.

If a customer books three times a year and stays two years, that's six bookings at $9.62, or about $58.

Against a $40 CAC, your LTV:CAC ratio is roughly 1.4 to 1. The commonly cited healthy target is 3:1 — you want to earn three dollars for every dollar spent acquiring, because the other two are funding everything that isn't acquisition. At 1.4:1 the channel isn't strictly losing money, but it isn't building a business either.

The related figure is CAC payback period: how long until a customer has repaid what you spent acquiring them. At three bookings a year, four bookings takes about sixteen months. That's sixteen months of funding the gap out of your own pocket. Ratios describe whether something works eventually; payback describes whether you can survive until it does.

Liquidity: the metric that's specific to you

Everything above applies to most consumer businesses. Liquidity is the one that's genuinely a marketplace concept, and it's the one investors probe hardest.

Liquidity is the probability that a given listing sells, or that a given search finds something worth booking. Depending on your model you'll measure it as fill rate, match rate, searches ending in a booking, or the share of active listings that transacted this month.

It matters because it's the leading indicator behind all the others. A marketplace with poor liquidity can buy growth on both sides indefinitely and still churn both sides just as fast — buyers who don't find what they want leave, sellers who don't get bookings leave, and your CAC has to keep replacing them. Fixing liquidity usually improves retention, LTV and CAC simultaneously, because it fixes the reason people were leaving.

Which of these to actually watch

Metrics divide fairly neatly into the ones that flatter you and the ones that don't. GMV, signups and traffic go up almost automatically as you spend more, which is what makes them comfortable and largely uninformative on their own.

The honest set is shorter: contribution margin per transaction, fully loaded CAC by channel, repeat rate, and liquidity. Those four will tell you whether the business works. GMV tells you how big it is, which is a different question.

Where referrals change the arithmetic

Run back through the example and notice which number was doing the damage. Contribution was $9.62 a booking, which is what it is. Repeat rate was three bookings a year, which is hard to move quickly. The number that made the ratio fail was the $40 CAC — and unlike the others, it was set by an ad auction rather than by you.

A referral programme inverts that. You decide what a converted customer is worth — a flat $10, or a fixed share of the revenue that customer generates — and that becomes your CAC. Not a forecast of it. It also only triggers on a completed transaction, so the people who clicked and didn't book cost you nothing.

At $10 a customer, the same $58 of lifetime value produces a ratio close to 6:1, and payback lands inside the second booking rather than the sixteenth month. Same marketplace, same margins, same customers. One input changed.

That's the useful thing about getting your metrics straight: it usually becomes obvious which single number is the one worth attacking.