When evaluating any business, it is important to understand how much value it creates and how much it captures. This is especially true of platform and marketplace businesses. While platforms that take off often create a lot of value, they “capture” that value at very different rates.
In this piece, I’ll cover just how much variance there is in the ability of different platforms to capture value, and the factors which I believe drive that “take rate”.
Marketplaces and Take Rates
One way to get a sense of the value capture ability of a marketplace is by measuring its take rate.
The take rate is essentially the average percentage of the value of third party transactions which is captured by the platform or marketplace as fees or commissions. Mathematically, take rates are calculated by dividing the net revenue generated by a platform or marketplace by the gross merchandise volume of transactions that net revenue was generated on.
Using information available in public filings and in the publically available rate cards, I’ve compiled the estimated effective take rates for a set of marketplaces and platforms below.
A few things to note:
I’ll use platforms and marketplaces interchangeably, but we’re largely talking about platforms that facilitate some kind of transaction.
For public companies, I’ve typically calculated their effective take rate on 3rd party transactions (i.e, not including first-party sales).
For private companies, I’ve mostly used their public-facing pricing and rate cards and using estimates where needed.
The most striking thing in the chart above is how wide the range is in the take rates, all the way from close to 0% (Craigslist) to up to 85% (Shutterstock’s highest marginal take rate with ~70% being their average).
Factors Driving the Take Rate
So given this big variance, is it possible to determine what the potential long term take rate is for a given marketplace? Or how much a new marketplace should charge?
While I haven’t been able to come up with a formulaic approach, I’ve been able to identify a few key factors which tend to influence the take rate a marketplace can charge, which I go into below.
1. Gross margin and nature of the product sold on the marketplace
One key determinant of take rates is what the marginal cost is to provide the product/service being sold on the platform relative to its price.
All things being equal, it is possible to charge a higher take rate on digital products than physical ones, since the marginal cost to provide them is ~0. Consider stock photos, apps, and ebooks, which tend to have fairly high take rates.
For services and physical goods, when a marketplace drives a sale for a seller, the seller on the marketplace still needs to incur COGS to produce the good (and deliver, etc). But taking a step further, for low gross margin products, the potential take rate one can charge is lower than that for higher gross margin products.
Take the example of a Farfetch vs Etsy within a given product line of apparel. Not all the difference is down to this, but given that Farfetch is a luxury brand marketplace, the brands on it can give up a higher take rate and still come out profitable compared to say a lower gross margin product in the same category on Etsy.
2. Concentration of suppliers (and buyers) on the marketplace
For a marketplace, the more fragmented the suppliers on the marketplace the better, since it reduces their bargaining power. If suppliers are very concentrated, then they can bargain/negotiate for lower rates or threaten to not make their supply available on the platform, which can be detrimental in a concentrated market.
The best example to highlight this is by digging into Expedia, which sells both flights and hotels. On the flight side, there are typically only ~4 airlines that matter for a given geography, so the supply is very concentrated. Meanwhile, on the hotel side, the supply is much more fragmented. While they don’t breakdown the take rate by segment, most estimates suggest that on the hotel side, their take rate is in the 15-20% range, while on the airlines side it is in the 3-5% range. The concentration is a key driver in this. Think about if Expedia didn’t have United when you performed a search. If you wanted the best price, you could search on Expedia, and then separately search on United before making your purchase.
Note that the concentration of buyers on the marketplace also matters for the same reason, but practically speaking a lot of the marketplaces highlighted above have consumers or SMBs as the buyer, which makes them naturally pretty fragmented.
3. Competition and alternatives
Naturally, one of the elements which determine the take rate platforms can charge is competition.
I generally think about “competition” in this case along two axes:
Have other providers have aggregated enough demand and supply in that category: Other things being equal, the more that exist, the lower the take rate will tend towards over time.
What is the best alternative to not using a marketplace: How painful or cumbersome is trying to do the transaction not on the marketplace and what is the typical take rate involved in that process. For example, in the offline world in many markets, brokers help facilitate transactions that
How easy is to disintermediate the marketplace over time: Typically when transactions are some combination of very high-value transactions, recurring in nature and involve the buyers and sellers meeting/communicating a lot, the alternative isn’t just to not use the marketplace but to use it to meet the other side and then go off-platform. In those cases, take rates should be lower.
More broadly, one can view the above as thinking about: what is the best next alternative to not using this specific marketplace?
One special case of this if this is no real alternative because the platform has a “monopoly” on a certain set of users. Take examples of the Apple App Store (or Google Play Store). For all practical purposes, the only way to provide in-app purchases or paid apps to the billions of users who have an iPhone or an Android phone is through these stores (sideloading apps on Android phones aside). This allows them to charge more than they might have been able to if there were better alternatives.
Note that they could arguably charge even more, but there are other considerations such as regulatory ones or wanting to maximize the total pie rather than the percentage share of the pie captured.
Another point here is that markets that have strong network effects and high switching costs and that are difficult to multi-home tend to lead to winner-take-most or winner-take-all outcomes. In those cases, the alternatives to the winning marketplace are obviously worse leading to the potential for higher take rates for the winner.
4. Ability of the marketplace to drive incremental distribution
One of the most critical ways a marketplace can provide value is to facilitate transactions for the seller which they believe might not have happened otherwise.
Marketplaces that have aggregated demand effectively and can drive additional (i.e., incremental) sales for the supply side on the marketplace can charge higher take rates.
One way to see this is to consider the difference between Shopify and Amazon. Shopify provides merchants the tools to set up a store and process transactions but isn’t necessarily bringing them sales and traffic. Meanwhile, Amazon has aggregated 100s of millions of buyers, and sellers lose the ability to reach them on a given search (and potentially make an incremental sale) if they aren’t on Amazon’s third-party marketplace. Shopify’s effective take rate is ~3% while Amazon’s is 10-15% depending on the category of product.
In fact, some marketplaces go as far as charging different take rates on transactions depending on their level of contribution to driving the transaction. Take Udemy’s pricing structure, which highlights the importance of distribution:
If the course is sold through the instructor’s link → instructor keeps 97%
If the course is sold organically on Udemy → instructor keeps 50%
If the course is sold through Udemy’s partners’ → instructor keeps 25%
Udemy charges a much higher take rate in cases where the sale occurred because of Udemy’s efforts in distribution.
5. Value created from Value-Added Services
Aside from just connecting buyers and sellers of some form, many marketplaces provide value-added services on top. Some of these are fairly ubiquitous across marketplace categories (payments) while others are either helpful or at times necessary in specific categories.
The benefit of some of these value-added services below is that aside from raising take rates, they can also lead to a more liquid and smoother functioning marketplace. However, it is important to note that there are costs involved with offering them, and so it’s important to consider the unit economics on each transaction including the costs, to get a better understanding.
People often use the term managed marketplaces to refer to marketplaces that have these value-added services, but just given how broad the term is, for clarity I’ve stayed away from the unmanaged vs managed distinction and instead touched on specific value-added services that can drive up the take rate.
Most marketplaces support payment processing and transactions on the platform. Just given the cost associated with payments, supporting this allows marketplaces to charge 3-5% additional.
Many marketplaces even specifically break out their take rates on their rate cards as X% + payment processing fees. Examples include Substack which charges 10% + payment fees, StockX which charges 7-9% + 3% payment fees, Kickstarter which charges 5% + 3-5% payment fees.
B. Quality Control and Authentication
For functional markets, trust is critical, and one way to build that trust is by doing quality control and authentication. Marketplaces that do this not only often have smoother functioning markets but also can charge higher take rates.
One example is GOAT vs eBay. You can buy sneakers on either, but GOAT has a 14% take rate vs 9% on eBay. One of the key reasons for this is that GOAT verifies the authenticity of sneakers, whereas eBay is pretty much unmanaged.
C. Insurance and Fraud Protection
Another value-added service that helps build trust and a smoother functioning ecosystem is providing insurance to one or both sides of the platform.
Examples include Airbnb which has a host protection guarantee of up to $1M and Uber, Lyft, and Doordash all provide insurance for drivers during a gig of up to $1M.
Anecdotally, especially on long stays, Airbnb’s take rate can be expensive and given communication between the buyer and seller, it is quite straightforward to go off-platform. However, Airbnb’s protection for both sides makes often means that both sides prefer to book through Airbnb despite having to pay Airbnb the ~12-15%.
Given the importance of incremental distribution, one of the best ways for a marketplace to grow its effective take rate is through advertising which involves essentially charging extra for additional distribution. This can probably be a post in itself, but the beauty of advertising is that it allows the suppliers who want more distribution to pay for it, while not increasing the cost of the platform for those who don’t want more distribution.
One example is Amazon which now has a ~$20B advertising business. While it charges suppliers 10-15% to sell on their eCommerce marketplace, those who want to do more volume can pay more to appear in search results, which increases the effective take rate Amazon gets from these merchants on these transactions.
In summary, the five factors I believe play a key role in determining the take rate:
Nature of product or service
The concentration of buyers and suppliers
Competition and alternatives
Ability to drive distribution
Value-added services provided
It should be noted that while the take rate is an important metric, the goal of a platform should not be to maximize its take rate. The goal typically should be to maximize total long-term profits generated, where profits = (take rate x average transaction value - cost per transaction) x (# of transactions)
This means a few things:
lower take rates in the short-term might be beneficial when facing a lot of competition or trying to grab share or get to scale in winner-takes-most or all markets.
there is often a trade-off between take-rates and the number of transactions, and so the lowering take rates may be beneficial for the marketplace overall.
increases in the take rate that come at the expense of even larger increases in cost per transaction may actually be detrimental unless also benefitting the marketplace in other ways (increased liquidity/trust).
Thanks for reading! If you liked this post, give it a heart up above to help others find it or share it with your friends.
If you have any comments or thoughts or if there are any corrections or additions to the chart, feel free to tweet at me.
If you’re not a subscriber, you can subscribe below. I generally post once a week or every other week on Monday about things related to technology or business.