How to Set and Test Your Marketplace Commission

Before choosing a commission percentage, write out a complete transaction: the buyer’s total, the supplier’s proceeds, your fee and the costs you carry. The rate is then a hypothesis to test with the people paying it.

Directorism designs and builds marketplace payment flows around explicit fee rules. A useful pricing decision connects the value customers receive with the costs, checkout experience and supplier proceeds behind the percentage. To gauge value, ask practical questions: What alternative does the supplier use today, and what does it cost them? How much time does a buyer spend finding this service without your platform? Use these to form a hypothesis, but remember that a user's stated willingness to pay does not guarantee their actual paying behavior.

Define the Fee Base Before the Percentage

Before stating a rate, define the exact base figure the percentage applies to. Clarify whether your fee applies to the supplier’s listed price, the buyer-paid total, or a figure that excludes variables like taxes, shipping, or tips, according to your disclosed contract.

Consider this hypothetical transaction, with all amounts in the same currency:

• Base price: 100 units
• Supplier fee: 10% of base (10 units)
• Buyer fee: 5% of base (5 units)

The buyer pays a total of 105 units before any separately excluded charges. The supplier receives gross proceeds of 90 units. The platform collects gross fees of 15 units.

Calculated against the base price, the fees represent 15%. Calculated against the total amount the buyer paid, the fees represent approximately 14.29%. The percentages are assumed inputs to show the arithmetic, not recommended rates. Direct costs and refunds will further reduce your platform's actual contribution. Keep supplier proceeds separate from platform revenue.

Gross merchandise value (GMV) records transaction value on your stated base; it is different from platform revenue. Platform revenue can include a fraction of GMV as fees, but also subscriptions or lead fees. Stripe's guide to key marketplace metrics details this distinction.

Your payment flow also determines where provider fees and refund costs land. For example, with destination charges on Stripe, processing fees are debited from the platform balance. Refunding a charge does not automatically reverse every supplier transfer or application fee, so you must verify your chosen flow and refund settings to understand your exposure.

Choosing Who Pays

You can charge the supplier, the buyer, or split the fee between both.

For deeper context on structuring your platform's financial mechanics, review our guide on marketplace business model selection.

Testing and Monitoring Your Pricing

Calculate the realized transaction-fee rate as retained transaction fees divided by GMV on the same stated basis. Fee credits, discounts and your treatment of refunded transactions can change that ratio. Keep unrelated subscription or lead revenue separate when comparing it with the advertised commission.

When testing a new pricing hypothesis, change one variable at a time. Compare similar user segments and time windows, recording uncertainty and potential network spillovers. If your platform has low transaction volume, you cannot rely on statistical A/B tests to prove a winning rate; use direct user interviews and small, reversible pilots instead.

Review both sides of the transaction. Track completed transactions, supplier availability, repeat use, fee credits, user complaints, and overall contribution margin. A higher fee per transaction can be offset by fewer completed transactions or higher support costs. For existing users, implement a clear notice plan that handles your actual agreements rather than a fixed grandfathering period. You can learn more about measuring these operational signals in our liquidity measurement guide resource.

Pricing Test Decision Record

Define the parameters before running the test. Treat the examples below as prompts, not recommended fee changes.

• Hypothesis: [What change are we testing, and which customer behavior or contribution result do we expect?]
• Target Cohort: [e.g., New buyers and suppliers in Geography X.]
• Disclosed Terms: [The fee base, payer, amount, refund rules and notice shown to affected users.]
• Comparison and Observation Window: [Which comparable segment or baseline, measured for how long?]
• Minimum Acceptable Result: [Set your own contribution and completion requirements before observing the result.]
• Owner: [Name/Role]
• Review Date: [Date]
• Stop/Revert Conditions: [Which result or customer problem causes a review, and how can the change be reversed?]

Next Step: Model Your Economics

To formalize your math, use our Business Model Unit Economics Calculator (Excel).

To enter the 100/10/5 arithmetic example from earlier: log one transaction, set the GMV base to 100, and set the combined gross fee fraction to 0.15. You must record all direct costs separately. Record the buyer and supplier split alongside your fee example; the workbook uses the combined rate to calculate gross commission.

Enter zero for inapplicable categories and leave unknown inputs blank. Keep acquisition calculations tied to the same customer group, or cohort: divide the cost of acquiring that group by its new paying customers to calculate customer acquisition cost (CAC). Enter monthly contribution per paying customer from that same group separately. The workbook uses it for a constant-contribution payback calculation and an assumed lifetime contribution scenario. It does not infer customer value by dividing total platform revenue by new customers.

If you are testing an idea before launch, use the marketplace validation checklist to record buyer and supplier commitments before treating your assumptions as demand.

Your next step: Write out one complete fee example for your platform, model its contribution in the workbook, and document your smallest viable pricing test using the decision record above.