_FINANCIAL-OPERATIONS/
Designing Commission Models
By Clément Salaün · Co-Founder & CTO
Marketplaces
Loading article content
Loading table of contents
Loading author
If you’re operating a platform, you know that finding the correct monetization strategy is not easy—but also critical to your business. There are a lot of ways to structure a commissions model, and it’s all too easy to get stuck in a static model like it’s the 1970s. Let’s explore different commission models, and how to optimize the various tradeoffs each model has.
As a platform, we have a delicate balancing act among these constraints:
Our commission model is our primary tool of choice for balancing these constraints. It acts as an optimizing function that reflects which of these constraints are priorities. A poor model will result in under-optimization of all of these constraints, resulting in lost opportunities and revenue. But a smart model will know when to sacrifice one constraint to optimize another (e.g. reducing take-rate to increase liquidity)—resulting in increased revenue.
When designing a commission model, we have a couple of levers at our disposal: Who we charge the commission to, and how we calculate the commission.
The first question is to ask: Who is paying the commission? It might be the buyer, the seller…or some combination of both. Let’s suppose we have a seller who has an item they are willing to sell for $70. We in turn desire a $30 commission on the sale of this item. The sale of this item (and resulting invoices) might be structured as such, if we take each box to represent $10.
One option is to charge the commission to the buyer. The buyer pays $100 for the item, and we pay out $70 to the seller, keeping $30 as our commission on that sale.
Another way to structure the commission model is to charge the commission entirely to the seller. The buyer pays $100 for the item, which is paid out to the seller. Later, we later charge the seller $30 as our commission on that sale, so the seller nets $70.
We can also split the commission between the buyer and the seller. For example, we might charge the seller a third of the commission and the buyer two-thirds. The buyer pays $100, and we keep $20 of that as commission, paying out $80 to the seller. We then later invoice the seller $10 for their share of the commission, so the seller nets $70 on the sale.
In all of these scenarios, our take-rate remains the same—$30. But the structure definitely has an impact on the other constraints. Some impacts can be predicted, others will have to be experimented with as they will be context dependent.
When we charge the commission to the buyer, we must either generate a separate invoice for the buyer, revealing our commission structure, or we must become the merchant of record for the sale, which exposes us to risk if the product is defective or needs to be returned.
When we charge the commission to the seller, we have to invoice the seller separately, which they might not be able to fully recoverable as expenses, depending on the region they are operating in.
Of course, there is also the question of how much commission to charge in the first place. There are three basic options: Fixed-fee, a percentage, or a step function.
On this calculation model, we charge a fixed free regardless of the sale price. This has the benefit of offering predictability to buyers and sellers when pricing is uncertain, for example because it has to be negotiated, or depends on larger market conditions. If the selling price goes down, you retain a good commission. But you lose out on commission if the price goes up.
On this model, we charge a percentage of either the public price of the product, or of the seller payout. When pricing is certain, our commission scales with the price. But if the price is too low we might find that the commission is not worthwhile. And if the price is too high, the seller might not see the value of our services compared to the commission we charge.
Of course, you can combine fixed fees and percentages in lots of interesting ways depending on the price point, in order to optimize our revenue. Here are some examples of how that might play out.
We might want to focus on increasing the breadth of lower-priced items in our catalog, because we have price-sensitive customers. We know that items priced under $100 have an 8% conversion rate, while items priced over have only a 4% conversion rate. So, to incentive sellers to increase their range of lower-cost items, we might drop our commission to 1% on the first $100 of an item’s price. Doing so should increase supply volume, liquidity, and average conversion rate, resulting in a greater take for us than setting a 10% commission across the board.
Or maybe it’s the opposite. We want to attract a higher-end segment with premium products. However, our commission is disincentivizing our sellers, because 20% of a $10,000 sale is not proportional to the value we bring the seller. Instead, we could align our incentives by charging 20% on the first $1,000 of an item’s price, and 5% on what remains over that. This will encourage sellers to focus on more expensive products.
One important consideration is how prices are defined and communicated with sellers. Both the payout price and the public price have an impact on the liquidity relative to each seller.
Clearly, each seller has an interest in their payout price being higher.
But they also have an interest in the public price as well. In some cases the manufacturer of a good sets that public price (MSRP, or Manufacturer’s Suggested Retail Price), or places caps on discounts. These restrictions are often enforced via the seller’s contract, and thus both we as the platform and the seller must respect.
More generally, if the seller is aware of the market rate for their products, they understand that when the public price is higher than that market rate, liquidity is going to decrease as potential buyers look elsewhere.
So, if we attempt to optimize our margin by increasing the public price, we risk putting certain sellers at a disadvantage—or risk losing those sellers altogether if we violate MSRP demands.
We have two options as a platform negotiating the price with a seller.
First, the seller may have a certain desired payout. In this case, we can compute our commission as a percentage of the public price in order to reach that desired payout. Or, we can compute the commission as a percentage of the desired payout. In either case, we might want to retain the right to increase the public price and either pass some of the additional earnings to the seller, or recover those earnings as part of a commission on the buyer.
Or, the the seller may desire a certain public price. In this case, we compute our commission as a percentage of the public price. We might still charge the buyer additional fees not factored into that public price.
As you can see, we have a lot of levers at our disposal for structuring our commissions to optimize the various competing constraints. No doubt, it will take experimentation to find the right commission model for our audience.
That said, if we’re going to experiment—and we should—there is no reason to settle on a static commission structure. Instead, we should consider dynamically altering our commission structure to reflect changing market conditions, a topic we will dive into in a future blog post.
