The case for user ownership of online marketplaces (part 1)
The web2 playbook for starting and scaling marketplaces
At the end of my first post where I wrote about web3 marketplaces’ growth flywheel, I said user acquisition is just the beginning of the magic when it comes to building in web3. I planned to make the case for why user ownership represents such a huge upgrade to the way we built online marketplaces in web2.
It seemed straightforward enough. Then, I started trying to actually write it, and it seemed like the number of ideas I had to make connections between in order to do this topic justice just exploded:
So consider yourself warned that I probably missed some obvious insight, misunderstood something I read, or am generally wrong in some other way in this post. Let’s just pretend like we’re two friends talking about this on a long walk, or maybe at a party. Since we’re just having a conversation here, please comment and / or get in touch if you have thoughts to share!
The web2 playbook for starting and scaling marketplaces
Online marketplaces have been some of tech’s biggest winners in the last decade, and Ben Thompson’s concept of aggregation theory really helped me understand the factors driving this success. To paraphrase one of the core insights from his work:
The internet unlocked a new abundance of ways for suppliers to connect with consumers. Before the internet, access to consumers was limited, with physical constraints like geography, space on billboards / printed things people see, and shelf-space contributing to this scarcity. But with the internet, anyone can market goods and services digitally to consumers anywhere in the world. The new scarcity became the limited time and attention consumers have to browse the sea of practically infinite options.
Even more concisely, businesses used to compete over access to consumers, but the internet made that access incredibly abundant, so now businesses compete over consumers’ attention.
Enter the business model that’s perfectly poised to take advantage of this shift: aggregation.
Aggregators win consumer attention by bringing together a lot of good choices conveniently in one place so the consumers don’t need to look anywhere else.
And this model works especially well in situations where browsing the supply for a given good or service is hard without an aggregator (e.g., the supply is very fragmented, is of inconsistent or indeterminant quality, requires trusting strangers on the internet which is scary, etc.). The more that large amounts of time and attention are required for consumers to navigate the supply for a given good or service, the more value an aggregator can potentially unlock.
To tie this to some real-world examples:
Airbnb made it easier to navigate the supply of hosts looking to rent out spare rooms and unused vacation homes by building a reputation system for trust between hosts and guests
Uber made it easier to navigate the supply of drivers willing to take people around town in their own cars and happens to be near the rider’s location at the time they need a ride
DoorDash made it easier to navigate not only the supply of restaurants offering food to-go, but also the supply of delivery drivers willing to pick up the food and take it where it’s supposed to go
Before each of these businesses existed, it was prohibitively difficult to find and transact with the highly fragmented and sometimes questionable supply of strangers in these respective markets. Online marketplaces create value by bringing people together - they attract both suppliers and consumers onto their platform and use technology to match them frictionlessly, at lower costs than was previously possible.
And because of the nature of aggregation, online marketplaces have powerful network effects. A virtuous cycle emerges where 1) more consumers on a platform lets it attract more suppliers, 2) more suppliers enhances the experience for consumers, attracting more consumers, 3) over time most of the suppliers and consumers for a given good or service end up coming together on the same platform, and 4) that platform emerges as a clear winner for being the one place that can connect the highest number of people with the lowest friction.
This dynamic means successful online marketplaces tend to win the market and become natural monopolies.
But before you start thinking building an online marketplace is easy…
Let’s take a look at Uber’s numbers leading up to it’s 2019 IPO:
Uber, one of the most successful online marketplaces ever created, lost a total of nearly $20 billion dollars in the four years between 2016-2019.
To understand why this happened, here’s a line from Andrew Chen’s excellent book on launching and scaling marketplaces, The Coldstart Problem:
If you have a chicken and egg problem, buy the chicken.
All marketplaces face a chicken and egg problem because they need to reach a critical mass of both supply and demand before they can start doing their magic.
Prior to reaching this point, marketplaces spend a lot of money in order to bring both sides of the market together, often over and over as they expand to new geographies or service categories.
For example:
Every time Uber launches in a new city where it has no drivers, no rider has any chance of successfully requesting a ride, so they won’t bother with using the Uber app until enough drivers sign up. Drivers, meanwhile, won’t get on the app either until enough riders are requesting rides. At the start, neither drivers nor riders get much value from the app. Uber solved this problem with money, in the form of driver subsidies (guaranteeing the minimum amount of money a driver could make by making themselves available for ride requests), referral credits (for both drivers and riders to invite their friends), and rider discounts (to bribe riders to change their behavior / switch over from professional transportation companies) - it essentially paid its early users to use its app in order to seed its initial network with enough supply and demand to get things going.
And Uber is not alone is using this “buy the chicken” strategy. Online marketplaces collectively spent so much money on acquiring initial supply and demand that tech journalists hilariously called it the millennial lifestyle subsidy. If you’ve gotten free groceries from Instacart, or dinner delivered on DoorDash for only a few dollars, you’ve been a beneficiary.
Every web2 marketplace startup does some version of this at some point in their lifecycle - it’s just what it takes to get the growth flywheel going. This spending is multiplied if there is more than one company trying to attract users in a competitive environment (e.g., Uber vs. Lyft, DoorDash vs. Grubhub, etc.): things tend to get orders of magnitude more expensive because there is pressure for companies to try to outspend their competitors. And to pay for all those subsidies that are going toward attracting users, marketplaces rely on investors to pick up the initial tab.
To sum up, here’s the web2 playbook for building online marketplaces:
Find a product or service where there is consumer demand, but navigating the available set of suppliers is costly due to consumers’ limited time and attention, which presents an opportunity to service that demand at lower cost
Aggregate enough suppliers on the platform, making it possible for the consumers to browse this supply at reduced time and attention costs
Aggregate enough consumers based on this value proposition, then start matching suppliers with consumers
Repeat, attracting as many suppliers and consumers as possible, ideally until it no longer makes sense for any suppliers or consumers to look anywhere else in order to find each other (i.e., the marketplace has won the market / achieved monopoly)
Raise and spend a lot of investor money along the way
So, building online marketplaces cost money, which means online marketplaces need investors.
Investors, of course, are only funding the expensive growth of online marketplaces because they expect these businesses to one day make that money back for them.
Unfortunately for everyone involved, no business can keep losing money on the margin and expect to make it up on volume. At some point, marketplaces have to flip the switch between spending investor money to fuel their growth and making money off of the users they have attracted.
Enter the “rake,” or the commission online marketplaces charge for providing the service. You can think of this as the platform’s profit margin, usually expressed as a % of a marketplace’s GMV (gross merchandise value), or the total value of goods or services that’s being paid for through the platform. The rake tends to start out negative (during the subsidy stage of a marketplace’s lifecycle), and increase overtime as the platform adds various fees in order to capture the value that it’s creating for users.
To understand how marketplaces decide when to flip this profitability switch, Bill Gurley has a very insightful post, A Rake Too Far: Optimal Platform Pricing Strategy, that highlights the strategic importance for a marketplace to optimize the balance between pursuing growth and profitability:
If your objective is to build a winner-take-all marketplace over a very long term, you want to build a platform that has the least amount of friction (both product and pricing). High rakes are a form of friction precisely because your rake becomes part of the landed price for the consumer. If you charge an excessive rake, the pricing of items in your marketplace are now unnaturally high (relative to anything outside your marketplace). In order for your platform to be the “definitive” place to transact, you want industry leading pricing – which is impossible if your rake is the de facto cause of excessive pricing. High rakes also create a natural impetus for suppliers to look elsewhere, which endangers sustainability.
Set the rake too high and growth would stop or become negative as users start going around the platform to avoid the fee, going to a competing platform that charges less, or just stop buying the good or service the platform offers altogether.
The virtuous cycle in the aggregator model where “more suppliers → more consumers → eventually everybody ends up on one platform” can quickly become a vicious one if things start moving in the opposite direction. Network effects are a double-edged sword.
Online marketplaces can be pretty brutal businesses to run, and they have investors to answer to, so this necessitates an equally brutal solution:
Returning to the point from the first section that “successful marketplaces tend to be natural monopolies,” the play here is to take over the market first, then increase the rake when users have nowhere else to go.
Imagine a negotiation between a marketplace business and its users:
The users want to keep the rake low, and their leverage comes from their ability to stop using the platform if they don’t like the deal being offered (e.g., the rake is too exorbitant).
The marketplace business wants to increase the rake, and its leverage comes from providing a service that is so valuable to users that users would rather pay more than lose access to the service.
The more a marketplace successfully aggregates both supply and demand, the more users lose their negotiating power, which allows the platform to get increasingly extractive with its rake without losing users:
When a marketplace enters the “extraction” stage of it’s lifecycle, the users who feel 1) the most trapped by their loss of negotiating power and 2) the most hurt by this extraction get very (understandably) upset. These users start to publicly protest against the platform, file lawsuits, and push governments to take antitrust action.
That means if an online marketplace manages to make it this far, it has to add one more step to its playbook:
6. Navigate adversarial relationship with the users it was supposed to serve.
The problem with this
The thing that I found the most counter-intuitive about all this is that an adversarial relationship hurts the marketplace just as it hurts users, since navigating this tension costs resources that could otherwise be spent on more growth or be returned to investors as profits. So unless we believe online marketplaces are run by sadistic jerks who want to fight their users, this is a situation that everyone wants to avoid. When all the value in a network is created by the users in it, doing something that upsets users should be the last thing the network wants.
So, there’s an apparent disconnect between “create value by bringing people together” and “a bunch of your users hate you.” It also means there’s an opportunity to do better.
More to come on what I think that opportunity might look like, starting with some thoughts in the next post on what I learned from the web2 playbook.
This has been part 1.
Coming up in a future post:
What works well about the web2 playbook, where it goes wrong
How user ownership in web3 upgrades the old playbook
How better playbook → better outcomes
Until then, thanks for reading!