50 Shades of Marketplaces
We started StubHub less than five years after eBay invented the low-touch marketplace but we diverged from eBay in that we chose a higher-touch model. We oversaw the fulfillment of the tickets through API integration with FedEx and handled all customer support when users had questions or issues to resolve. We guaranteed every transaction which brought peace of mind to our users and made feedback obsolete. All payments flowed through our bank accounts so we could provide refunds if necessary. We offered these value-add components because we knew it would lead to a more seamless user experience, more loyal users, and a 10x improvement over the existing options provided by eBay and Craigslist. Many people told us we were taking on too much operational burden, our margins would suffer, and we wouldn’t be able to scale.
Fast forward 20 years to today. We have seen several recent public debuts of world-class marketplaces including Uber, Airbnb, and Doordash that have diverged even more from the original low-touch marketplace that eBay invented over 25 years ago. eBay’s low-touch marketplace model was simple and efficient. Today’s marketplaces have chosen to take on more operational intensity, and thus lower gross margins, in the interest of providing a seamless end-to-end user experience. This is a smart tradeoff that has powered the growth and success of many modern marketplaces. However, some entrepreneurs have prioritized user experience to such an extent and taken ownership of so many operational functions that they can no longer claim to be a marketplace.
Classical low-touch marketplaces
eBay’s original marketplace model was highly scalable, required minimal levels of human effort to grow after reaching escape velocity, and operated in a truly capital-efficient manner. This low-touch marketplace model was highly efficient because it consisted primarily of software without significant operations; most transactions did not require any interaction between employees and users. Entrepreneurs who followed this template could simply build a web app, spend some money on marketing to acquire initial supply and demand, watch transactions start flowing with little or no involvement from marketplace employees, grow organically from word-of-mouth and snowball dynamics, accelerate growth with additional marketing spend, gain momentum as the flywheel spun, and deliver gross margins above 70%.
High-touch marketplaces (or managed marketplaces)
Many of today’s marketplaces take responsibility for more operational functions and are thus higher-touch businesses. For example, when a product marketplace is responsible for the warehousing, logistics, and fulfillment (even if those tasks are subcontracted to a third party) it is no longer a low-touch marketplace. Many people use the term “managed marketplace” to describe these models, though I prefer the high-touch nomenclature. While high-touch marketplaces generally have lower margins, they can still be amazing businesses so long as they have powerful network effects that attract more demand when there is more supply and vice versa. It is important to keep in mind that the distinction between low-touch and high-touch marketplaces is a spectrum with shades of gray and not a black/white bifurcation.
“The nirvana that investors want is a low-touch marketplace with a seamless user experience, but, alas, those are hard to find.”
Like StubHub, most high-touch marketplaces came into being because entrepreneurs wanted to ensure consistency and delight users. This type of consistency is much easier to deliver if you control key parts of the experience. As another example, compare Amazon (a high-touch marketplace that handles fulfillment for most transactions even though well over 50% of the products are sold by 3rd party sellers) to eBay (a low-touch marketplace with very little transaction-level involvement from eBay employees). Amazon is a much larger business with a more seamless user experience even though its gross margins are lower. The nirvana that investors want is a low-touch marketplace with a seamless user experience, but, alas, those are hard to find.
At Craft, we are very comfortable investing in both low-touch and high-touch marketplaces. And like the Amazon/eBay example, we know there are great investments in high-touch marketplaces even if the margins are lower.
The marketplace spectrum
Where does your marketplace fall on the spectrum between low-touch and high-touch? The key areas to think about when answering this question are in the table below. Most companies will have a mix of both low-touch and high-touch characteristics.
Sometimes entrepreneurs take on so much operational intensity or lack certain key elements that they cease to be a marketplace at all. Here are some examples of businesses that generally should not claim to be marketplaces.
- Businesses that take inventory risk
- Businesses where the supply side is made up of permanent W2 employees
- Businesses that charge a fixed fee for access to software (ie, a fee that does not depend on the dollar value or number of transactions)
- Businesses that just connect supply and demand but do not control payments
- Businesses that process payments but don’t generate incremental demand
If your business isn’t a marketplace, then what is it? Here are some guidelines to figuring yourself out:
If your company takes inventory, it’s not a marketplace. Marketplaces have no capital tied up in inventory and no inventory risk. Inventory risk refers to the risk that a business can’t sell the inventory it has and also has the risk associated with returned items. In ecommerce and retail businesses, this leads to markdowns to move out the unsold merchandise. Many ecommerce retailers offer products from third-party suppliers. But because they take inventory in those products, they have inventory risk and are thus not marketplaces. They are just retailers.
Tech-Enabled Service Providers
In a labor marketplace, the analog to “inventory risk” is having full-time permanent W2 employees. If your business has full-time employees who provide the services to the demand side, you have “inventory risk” that the demand will not materialize to pay for your employees and you are really a tech-enabled service provider. Even if your business uses 1099 contractors but you are negotiating long-term prices with the demand side, you’re somewhere in between a tech-enabled service provider and a high-touch marketplace.
Marketplaces have revenue that is directly proportional to the GTV. If your business offers SaaS software that enables users to transact with others but your pricing is based on a fixed SaaS fee, you are a SaaS business, not a marketplace. One common and very effective technique of building a marketplace and solving the “chicken and egg” issue is to build a SaaS tool that is useful for one side of a marketplace. After enough users are relying on the SaaS tool, the business can introduce the marketplace functionality and start marketing to attract the other side. But if you employ this strategy, you are just a SaaS business until you have both sides transacting and a take rate revenue model.
If your product enables companies to process payments from their existing customers but does not generate incremental demand from new customers, you have a payments business, not a marketplace. Like marketplace take rates, payments businesses benefit from generating revenue that is proportional to total spend. However, payments businesses typically only generate 1–5% net payments revenue (after the variable cost of processing the payments) while marketplaces often command a 10–20% take rate. The higher take rate is justified by the fact that marketplaces deliver new demand from new customers to the supply side.
If your business doesn’t control the payment flows, then you are more like a classified listing service. Craigslist is the original example, but there have been others such as OLX, and eBay’s Classifieds business. Many of these companies have been wildly successful and very profitable because there is almost no operational intensity to these businesses (eg they are so low-touch that they fail to qualify as a marketplace). However, they leave something to be desired when it comes to user experience and customer satisfaction.
Classical low-touch marketplaces are the most likely to grow on their own after getting the initial flywheel spinning, the most capital-efficient, and the most scalable. Despite these advantages, most opportunities we see today are high-touch marketplaces. Given the growing variety of marketplace models, there has been increased confusion about what exactly a marketplace is and which businesses qualify. We hope this post helps provide the answer.