How to Price a Two-Sided Marketplace: Take Rate, Subsidies, and Who Pays First

“Should we charge homeowners, should we charge contractors, or should we charge both, and how much” is the question that stalls more marketplace pricing decisions than any other, and staring at competitor pricing pages doesn’t answer it, because competitor pricing pages don’t show the thing that actually determines the right number — which side of the marketplace can least afford to leave.

That’s the real question behind marketplace pricing, and most founders and PMs skip straight past it to argue about take rate percentages instead. Get the “who pays first” question wrong and no take rate saves it, because the fee structure ends up optimized on a marketplace that never reached liquidity in the first place.

These debates almost always start the same way: someone pulls up a spreadsheet of competitor take rates, as if the right number is a market average to benchmark into. It isn’t. Two marketplaces in the same category can sustain wildly different take rates depending on how much real value they add beyond matching supply and demand — and copying a competitor’s number without copying the thing that earns it is how teams end up pricing themselves out of the liquidity they haven’t built yet.

Why Marketplace Pricing Isn’t Just “Pick a Take Rate”

Single-sided SaaS pricing is comparatively simple: one customer, one willingness-to-pay curve, optimize against it. A two-sided marketplace has two willingness-to-pay curves that interact with each other, and pricing one side changes the value proposition for the other side before a single dollar is collected. Charge contractors too much to list jobs and supply thins out, which makes the marketplace worse for homeowners, which reduces demand, which makes supply even less willing to pay — a genuine downward spiral that a single-sided pricing model would never produce.

This is why the foundational academic treatment of two-sided markets frames pricing as a joint decision, not two separate ones: the price on each side has to be set with explicit awareness of the cross-side network effect it triggers, not in isolation (Eisenmann, Parker & Van Alstyne, “Strategies for Two-Sided Markets,” Harvard Business Review, 2006). Nearly two decades later that framing still holds up better than most of the marketplace advice circulating today, because most of that advice quietly assumes liquidity already exists and skips the harder problem of how to get there.

Who Pays First: Subsidizing the Side With More Elasticity

The practical version of “who pays first” comes down to elasticity: which side will walk away from a fee faster. In most marketplaces, that’s the supply side early on — contractors, drivers, sellers, hosts — because they have other channels to get customers and no obligation to try a new one. Demand-side users, by contrast, are often price-insensitive relative to the value of finding what they need in one place, especially early, when few alternatives aggregate the same supply.

That asymmetry is why most successful marketplaces subsidize supply and monetize demand early, even when the eventual steady-state fee structure looks different — a pattern Reforge’s marketplace research documents directly, framing early subsidy not as a discount but as the cost of building the liquidity that makes the marketplace worth paying for at all (Reforge, “Understanding Marketplace Liquidity”). The resistance to this is predictable: supply is the harder side to recruit, so intuition says charge the side that’s easier to get. That intuition runs exactly backwards for liquidity purposes — the side that’s harder to recruit is the side worth protecting, because losing them costs the whole marketplace, while losing an easy-to-acquire demand user costs one transaction.

A recurring pattern: a marketplace charging both sellers and buyers a fee from day one, reasoning it’s “fairer,” watches seller supply stay thin for months because a listing fee isn’t worth it for anyone testing the platform casually — and thin supply means buyers have little reason to return. Dropping the seller fee to zero and moving the take rate onto the buyer side typically increases seller signups and buyer retention together, since buyers who keep coming back find more inventory each time.

Setting a Take Rate Without Killing Liquidity

Once who pays is settled, the take rate itself is a narrower question, but it’s still not “what can the market bear” in isolation — it has to account for what the marketplace actually adds versus a direct transaction between the two sides. A marketplace that only aggregates listings without adding trust, logistics, payments, or dispute resolution has a hard ceiling on what it can charge before both sides start disintermediating around it, arranging the deal off-platform after the marketplace did the work of connecting them.

The marketplaces that sustain higher take rates — 15–20%+ in some services categories versus 3–8% in commodity goods categories — are the ones where the platform is doing real work beyond matching: payment processing, insurance, quality guarantees, dispute resolution, logistics coordination. A marketplace with a thin value-add should price accordingly, or invest in making the value-add real before raising the take rate. This is also where a pricing strategy document earns its keep — the place to write down explicitly what’s being charged for beyond matchmaking, so an increase can be justified against added value instead of just announced.

This also connects to a supply-strategy choice: whether the marketplace is running a comprehensive, exclusive, or curated supply model, since each implies a different amount of platform work and therefore a different defensible take rate (Reforge, “Marketplace Supply Strategy”). A commodity goods marketplace — used furniture, secondhand electronics — sustains a low take rate because the buyer and seller could plausibly find each other another way; the real value-add is mostly convenience and reach. A services marketplace handling a $12,000 kitchen renovation sustains a much higher take rate because the platform is underwriting real risk — vetting the contractor, holding payment in escrow, mediating disputes — and that risk-bearing function is worth a meaningfully larger cut than pure matchmaking ever will be. Pricing a commodity marketplace like it’s underwriting risk it doesn’t actually underwrite is one of the fastest ways to push transactions off-platform.

When Should You Charge Both Sides of the Marketplace?

Charging both sides makes sense once liquidity is established and both sides have meaningfully more to lose by leaving than they did at launch — once switching costs exist that didn’t exist during the cold-start phase. A ride-hailing marketplace that only charged drivers early on can introduce a small rider service fee once riders have enough trip history and saved payment methods that a competing app feels like real friction, not just a different icon on their home screen.

The mistake is introducing a second-side fee too early, before that switching cost exists, which just reopens the disintermediation risk months of work went into closing. Sustained week-over-week retention on both sides — not just total volume — is the signal that a fee on the previously-free side is worth testing, and testing it on a small cohort first matters, because marketplace users notice fee changes fast and talk to each other about it.

When Marketplace Pricing Breaks

Marketplace pricing fails in a specific way that doesn’t show up in the take-rate spreadsheet: the platform’s own fee structure quietly incentivizes the exact off-platform behavior it’s supposed to prevent. A freelance-services marketplace that raises its take rate from 10% to 20% without adjusting anything else about the value it delivers commonly sees repeat transactions between the same buyer-seller pairs move off-platform within a couple of months — at 20%, neither side feels like they still need the marketplace for a second transaction it already facilitated once.

A second common failure is treating take rate as a single global number when the marketplace actually spans sub-categories with wildly different margins — a home-services marketplace charging the same 15% on a $40 lawn-mowing job and a $12,000 renovation is over-pricing the small jobs and under-pricing the large ones. A third is timing: introducing a second-side fee right after a growth milestone instead of a retention milestone. Volume isn’t the right signal — a spike in transactions can be driven by a cohort that hasn’t yet formed a habit, and charging them before they have is often what keeps the habit from forming at all.

Worked Example: Pricing at a Series A Logistics Marketplace

A Series A freight marketplace connecting small shippers with independent truckers faced a real constraint: six months of runway and pressure to get revenue moving before the next raise, pushing toward monetizing both sides immediately rather than subsidizing supply the way the textbook approach would recommend.

The resolution split the difference. Truckers — the harder-to-recruit, more elastic side — paid zero listing or matching fees, preserving the subsidy that kept supply growing. Shippers paid a transparent booking fee on top of the freight rate, positioned as covering payment guarantee and dispute resolution rather than a marketplace tax — shippers who understood what the fee bought them complained far less than the ones who saw it as a pure markup. Weekly active truckers and repeat shipper bookings, not gross bookings, served as the leading liquidity signal.

Within five months the marketplace reached enough two-sided retention to test a small, optional trucker-side fee for premium load matching, adding revenue without reopening the supply risk a mandatory fee would have created. The slower path carried real runway pressure throughout — but marketplaces that charge both sides from day one under similar funding pressure tend to burn through supply and never recover enough liquidity to raise a Series B.

Marketplace Pricing Decisions, in Order

Question	What to Decide First	Why It Comes First
Which side is more elastic?	Identify the side likelier to leave over a fee	Determines who gets subsidized early
What are we charging for?	Matching only, or matching plus trust/logistics/payments	Sets the ceiling on a defensible take rate
Is liquidity established?	Check two-sided retention, not just volume	Signals when a second-side fee is safe to test
Does take rate vary by category?	Segment by transaction size and margin	Prevents a blended rate from mispricing both ends
How is the fee positioned?	Frame it as value delivered, not a platform tax	Reduces disintermediation risk after the fee lands

Make the Call This Week

Anyone staring at a marketplace pricing decision right now shouldn’t start with the take rate number. Start by writing down, honestly, which side of the marketplace would leave fastest if charged today — and price that side at or near zero until retention data says otherwise. The take rate percentage is the easy part once the harder question is settled.

References

• Eisenmann, Parker & Van Alstyne — “Strategies for Two-Sided Markets,” Harvard Business Review, 2006 — https://hbr.org/2006/10/strategies-for-two-sided-markets
• Reforge — “Understanding Marketplace Liquidity” — https://www.reforge.com/guides/understanding-marketplace-liquidity
• Reforge — “Marketplace Supply Strategy: Comprehensive, Exclusive, or Curated” — https://www.reforge.com/blog/marketplace-supply-strategy