👋🏻Welcome back, especially to my newest subscribers after my latest piece ‘Why DAO Contribution’ sucks. If you like reading in-depth essays that are actually worth your time, please subscribe by clicking below:
2021’s NFT market felt like a casino: Mints felt like a spin of the roulette wheel, Higher-than-floor listings like a bluff and Twitter threads like a dubious system designed to give you an edge.
NFT marketplaces seemed like casinos, taking their 2.5% cut on each irresponsible financial decision.
In gambling, the house always wins. In 2021’s NFT craze, NFT marketplaces certainly did.
But NFT marketplaces aren’t casinos. Yes, they benefitted from millions of folks basically gambling on JPEGs, but they didn’t make the odds or design rigged games.
In early 2023, the casino times are over. NFT volume has cratered, floor prices tanked. The few remaining traders look more like unwitting pedestrians gawking at a three card monte than cigar smoking tuxedo wearers enjoying a game of poker.
The few holdouts perform all this on a stage of economic turmoil in which markets are subtracting speculation from organic market volume. For the Web3 market, the result of this Great Subtraction will be the organic demand for NFTs (which hopefully exceeds zero).
We’re discovering which NFTs people actually want to own.
The end of the everything-goes-10x-because-JPEG-funny extravaganza challenges NFT marketplaces. Taking a cut of transactions is less lucrative when transactions are both fewer and smaller. For some marketplaces, this threat is existential. Are they taking the correct steps?
A good portion of what’s left of web3 looks like a cargo cult trying to summon The Next Bull Run™.
They seem to hope that if they just tweet “it’s time to build” enough (wagmi now sounds unreasonable to even the most fanatical web3 enthusiasts), hopefully before the VC dollars run out, a new speculative bubble will emerge and make them all rich.
This obsession with The Next Bull Run as Web3’s messianic savior feels like long-term ponzinomics. Don’t get me wrong: The next bull run may very well create wealth, innovation and progress. But the promise of a speculative bubble can’t be the raison d’etre for an industry.
If an enterprise—NFT or not—can’t maintain non-zero cash flows and equity values without a bubble, it shouldn’t exist. Web3 can only prevail if we create products and services people pay for—which creates businesses whose survival doesn’t depend on VC cash infusions.
I believe this is possible. And the same way the ICO bubble gave us Aave, some survivors of the 2021/2022 web3 hype can likewise become robust (if not antifragile) infrastructure of our digital economy. But that requires adhering to business fundamentals and creating useful companies from first principles.
Because they’re so fundamental, NFT marketplaces are a useful case study to form predictions of the web3 ecosystem as a whole and of individual players specifically.
This essay explores the future of NFT marketplaces from first principles, draws parallels to survivors of the dotcom crash and analyzes the viability of different business models.
Let’s start by understanding marketplaces and their competitive dynamics from first principles.
We can only understand NFT marketplaces deeply when we understand the principles of marketplaces (the same reason we started with transaction costs when analyzing DAOs).
The next few paragraphs are economics 101, but bear with me. At its core, a marketplace matches supply to demand via a price mechanism. This results in each marketplace requiring two parties to complete a transaction and generate revenue:
A buyer who wants to obtain something at the lowest possible price (demand)
A seller who wants to sell something at the highest possible price (supply)
Prices mitigate these competing incentives: They rise when demand exceeds supply and fall when the opposite happens.
The product experience thus improves:
For buyers: The more sellers there are
For sellers: The more buyers there are
This leads to something called network effects. Per Hamilton Helmer, network economies happen in “a business in which the value realized by a customer increases as the installed base increases”.
This makes the success of marketplaces self-perpetuating. The more users you have, the better the experience, the more you’re likely to attract. That’s why some of Silicon Valley’s biggest winners have been marketplaces: AirBnB, Amazon, Etsy, Uber, etc.
Network effects also work the opposite way: Fewer users create a worse product experience, which negates any great UX or novel features your product may have. Marketplace competition usually follows winner-takes-most dynamics.
Marketplaces, in other words, compete for liquidity, which measures how reliably someone can complete a desired transaction at their desired price. In Finding Power, Every’s Nathan Baschez calls this “The basis of competition”:
“The basis of competition is the collection of product attributes that cause customers to choose one product over another.”
Holding other factors constant, liquidity determines which marketplace builds network effects and wins. In NFT-land, OpenSea has captured the most liquidity (read: network effect) and operated as a quasi-monopoly for much of 2021 and 2022.
But as we’ve seen with the rise of Blur, network effects don’t cement a marketplace’s dominance. Competitive dynamics in network effects-driven industries can shift.
As per Nathan Baschez’ Why Winners Sometimes Take All, And Sometimes Don’t, early winners may decline for one of two (related) reasons:
TikTok is an example of this: Network-effects-powered Instagram thought people cared about their friends’ content, but TikTok realized people wanted entertainment.
The government intervention of AT&T is a stark example of external factors affecting a champion’s market dominance.
The NFT marketplace space is currently undergoing an assault of external factors right now, which makes OpenSea’s juggernaut position more assailable than ever.
Besides illegible-to-me factors like the financial health of the companies themselves, a giant contributor to transformation is less visible now, but will become apparent in the near future. It’s what I call the primary transaction thesis.
Many Twitter users enthusiastically predict the future of NFT trading. While their predictions are usually about prices or types of NFTs being traded, they all seem to assume NFT trading will mostly look like it did in ‘21/’22.
I believe NFT trading is about to transform based on a simple insight. This is the most important assumption this essay rests on: Most NFTs won’t go up in value.
You’ll realize this if you imagine a world after The Great Subtraction of speculation demand from real demand. If NFTs succeed as a product category, it won’t be because of speculation or trading. NFTs will succeed if people demand the experiences they unlock and the art they’re linked to. While 2021’s NFT market looked like a WallStreetBets field trip to the Ethereum blockchain, NFTs aren’t primarily financial instruments.
They’re digital items that could represent art, a software subscription, a membership, a ticket, an in-game item, a collectible… the list goes on. Pre-blockchain, most of those things were not investments, but purchases. I don’t see why that should change. Yes, the lack of physical logistics makes NFTs easier to trade, but that in and of itself won’t transform things into investment assets.
That doesn’t mean some NFTs won’t multiply in price and that some people won’t profit on them.
But speculation can’t remain the main use case of NFTs. Like physical things, almost all NFTs will decline in price after we buy them. Some specialized art dealers make money by speculating on art, but that doesn’t mean buying up the inventory at your local art gallery is a sound financial decision.
If NFTs are to be viable, we need them to unlock unique experiences people want for their own sake. This will mainly happen through primary transactions, not the secondary sales most marketplaces rely on.
Why? Creators/companies need money to sustain operations.
Syrupy fantasies of sustaining companies and artists through royalties only are just that: Syrupy fantasies.
Yes, royalties provide some income, but their economics suck: If you set a (high) 10% royalty and sell out a collection at a 0.1 ETH mint price, you’d need the price to 10x just to make one more 0.1 ETH. If your initial mint was 1000, you’d need every piece to change hands to get your day one cash flow. That’s just to stagnate. It gets worse if you want to grow. And if you want people to hold, the picture gets even more bleak.
For all but the top 0.1% of creators, royalties are a bonus, not a business.
There may be a moral case for royalties, but the economic one is weak. Compare mint proceeds to royalties & OpenSea fees:
Creatoers and companies earn more from mints than from secondary sales. Excluding outlier collections, it has been this way and it will continue this way. Most money will be made on primary transactions, the sale of an NFT from a company/creator to a customer. This creates multiple advantages:
If/when primary transactions do become the bigger driver of NFT volume than secondaries, you capture a slice of a bigger pie.
Peer-to-peer sales are unpredictable while primary sales let you control the price.
That said, let’s make a bold comparison:
eBay used to rule eCommerce. It started by facilitating used (read: secondary) collectible sales between amateur sellers and captured a network effect. eBay became the high-liquidity marketplace, which made it the winner—for some time.
When you think of eBay today, you probably think of a peripheral eCommerce player, not the undisputed champion. Like Britain after losing its empire, eBay is still around and functioning. But it languished as a shell of its former self, playing third fiddle to Amazon and Shopify.
eBay has not only declined, it has also abandoned its roots of being a second-hand collectibles marketplace. 80% of transactions are now of new goods sold by professional sellers. Within our context, this is the Primary Transaction Thesis at work: A shift towards primary transactions.
(Interestingly, eBay was founded on a vision of ‘the perfect market’ by a techno-optimist libertarian, which resembles the dogma of web3, which often stands in the way of mainstream adoption)
A peer-to-peer secondary marketplace under competitive pressure became a primary marketplace between companies and customers.
I believe OpenSea is in a similar spot to eBay: A functioning secondary marketplace has captured a lion’s share of market share, but reckons with deteriorating conditions and changing customer behavior.
To be clear: It’s not bad to be eBay (or Web3’s eBay). The company has a $24B market cap, founder Pierre Omidyar is estimated to be worth $20B and it employs around 10,000 people.
The triumphant winner of ecommerce, of course, is Amazon.
With the PTT in mind, Amazon’s rise seems logical: Amazon focused on primary transactions, which is simply a bigger market.
This reminds me of Nathan Baschez’ point about misunderstanding: While eBay reigned supreme in eCommerce, Amazon redefined the market. Amazon likely couldn’t have defeated eBay by trying to win at their game. Instead, Amazon’s rise was driven by an insight that online markets might mirror offline markets—where primary transactions reign supreme.
(This, of course, is post-hoc reasoning. But whether conceptualized as such by Bezos or analyzed in hindsight, the insight is useful for drawing further conclusions about the future of NFT marketplaces).
As this crystallized, the eCommerce market was redefined into one where eBay was the follower, not the leader—which allowed Amazon to capture the liquidity and win eCommerce.
I’m sure the Amazon/eBay rivalry also involves other factors. But eBay’s position maps to OpenSea’s—and questions its future dominance which creates space for someone to win in the NFT space.
Who’s best positioned to do so? Let’s survey the current landscape:
OpenSea is to NFT marketplaces what Google is to search engines. Yes there are alternatives, yes some people use them, but to most, they’re synonymous with the category.
With market share towering above the rest and a $13B+ valuation, it was first to get significant network effects and won the market.
But the market hasn’t stood still. Many competitors and other business models have sprung up. We’ll dive into their features and competitive positions in the next few paragraphs:
This is the OpenSea model. Users include OpenSea (duh), LooksRare, X2Y2, Solanart, Objkt, etc. You can trade any NFT unless it’s banned. Listing and buying are core mechanics, with some quality-of-life features like statistics, multi-buy, etc.
They make money when a listing on their platform is purchased. OpenSea is the leader, but there are others.
While OpenSea and others have shipped launchpads to facilitate primary transactions, these initiatives look more like experiments than anything else.
LooksRare and X2Y2 (and probably others) are marketplaces with similar features and business models OpenSea, except for one difference: Token rewards.
LooksRare’s $LOOKS airdrop was a shrewd strategic move to get users form OpenSea: It gave OpenSea users free tokens in exchange for using LooksRare.
The ‘tokenomics’ expanded: There are buying rewards, listing rewards, participation in revenue shares, etc.
These tokens promise decentralization with a “by the people, for the people” rhetoric. While these increased trading volume for a while, these tokens were usually treated as ‘free money’ and have declined 95%+ in price. They also attracted wash trades, where traders sell items to themselves with another wallet in order to capture the trading fee.
Token-enabled marketplaces have captured and retained some trading activity, but both X2Y2 and LooksRare remain peripheral players compared to OpenSea, even as they’ve lowered fees.
Whether you view token rewards as a valiant attempt at decentralization or a glorified cashback scheme, the frothy exuberance after the airdrop doesn’t seem to translate to long-term market disruption.
Reserving moral judgment, I believe token rewards are a competitive disadvantage for undifferentiated products.
It’s hard for an organization with less capital, smaller market share and essentially the same product can win against a first-mover in a network-effects driven industry. I don’t see how incurring expenses for paying people to use your product fixes this.
Sidenote: Blur may be a counterpoint here. While their token bonanza isn’t over yet, their growth has been spectactular—this is in part because the user experience is differentiated, while X2Y2’s and LooksRare’s wasn’t.
NFT Protocols are perhaps the most uniquely Web3 business model. These provide the basics of NFT marketplace infrastructure:
Protocols disintermediate the front-end of an NFT marketplace from its backend.
While some protocols have their own dedicated front-end (e.g. Zora.co), the core mechanism is similar: You provide the lego blocks and convince others to build something with them.
This makes it easier to build trading and liquidity into another application (because you can import listings and the like).
While the obvious business model is to charge fees for transactions, there are more speculative ideas: Zora’s Jacob Horne describes hyperstructures; free-to-use protocols that are valuable despite charging no fees because they’re such an important piece of infrastructure.
The success of a protocol, again, rests on its ability to become defensible and capture network effects.
If a product creator can improve their own user experience by importing liquidity via a protocol (which improves the experience of other protocol users, which makes the protocol more attractive for builders…), a flywheel kicks off.
But builders also import dependencies. If protocols “turn on the fee switch” or prioritize their own front-ends, building on protocols can become a liability.
So far, most NFT protocols came from existing marketplaces (e.g. OpenSea’s Seaport and Rarible Protocol), likely because there’s some existing liquidity and building from zero is hard.
When gas fees were high and a plurality of marketplaces made it hard to get the best prices, aggregators came along.
Aggregators like Genie and Gem (acquired by Uniswap and OpenSea, respectively) display listings from multiple marketplaces and let you purchase them in one transaction.
By reducing into one transaction, aggregators save users gas fees and help them find good prices without navigating multiple marketplaces.
They have a useful place in the ecosystem because they benefit both sides:
Marketplaces get demand funneled towards their listings
Buyers get cheaper prices and lower fees
Theres precedent for this in web2. eCommerce aggregators like price.com do the same thing for eCommerce sites. While web2 aggregators send users to other websites, smart contracts enable NFT aggregators to execute transactions on their own website. That way, a marketplace’s specific front-end matters less as long as they have the lowest prices (something Blur seems aware of).
Aggregators fill a need. They gained adoption quickly and two big ones were acquired quickly. But they also currently have no revenue model: a of them charge fees and affiliate commissions are rare in the NFT world.
This is a solvable problem: Marketplaces could launch referral schemes to reward aggregators, as major web2 e-commerce players have done. Aggregators could also charge users a fee on transactions.
(A PTT-esque recognition of the importance of primary transactions has also led to mint aggregators like mint.fun)
Aggregated marketplaces like Rarible and Blur combine a marketplace and aggregation. You can buy from all marketplaces on their platform, but you can also list natively.
This solves the business model problem by charging OpenSea’s transaction fees on secondary sales, but also introduces new principal-agent problems.
Pure aggregators and pure marketplaces are symbiotic. Marketplaces make money when aggregators help them process transactions and aggregators offer a better experience the more listings there are on marketplaces.
Aggregated marketplaces blur the line between symbiosis and parasitism.
Like other marketplaces, the goal of aggregated marketplaces is to capture liquidity. To do so, they need initial liquidity, which they get from aggregating orders.
OpenSea and other aggregation targets make money from aggregated marketplaces (because they process their orders). Because both compete for native listings, they’re also competitors. But once liquidity shifts toward an aggregated marketplace, normal marketplaces are incentivized to shut down API access, leaving aggregated marketplaces with their native listings only.
This puts aggregated marketplaces in a precarious position: Their value proposition (get the best prices on the market) might get subverted by their own success.
(this is not theory, it’s happening)
Most marketplaces we’ve discussed are for more or less everyone. OpenSea, LooksRare or X2Y2 aren’t optimized for a specific use case or target market.
Niche marketplaces like SuperRare and Foundation serve more specific audiences (in their case, artists & art collectors). They use the same technologies—a blockchain base layer, smart contracts that enable bids, offers and listings, etc.
But these are best defined by what they abstain from rather than what they do.
SuperRare’s UI reveals their focus on art. It uses words like “collecting” instead of “trading”, “artists” instead of “creators” and call their fee a “gallery fee”. Artwork consumes much of the screen and is the clear focus. There are no analytics, price movements or other elements for traders.
Compare SuperRare’s item page to OpenSea’s, where artwork and numbers are more balanced. Or compare it to Blur’s, which looks like a Bloomberg terminal.
(While Blur’s business model differs from SuperRare’s, it’s also a niche marketplace because it specifically aims at traders)
SuperRare only allows artwork minted on their site to trade on SuperRare—and only allow approved artists to mint. SuperRare deliberately misses out on earning commissions on big collections like Bored Apes, CryptoPunks or Doodles.
Because it serves the artist/collector segment better than anyone, it can charge higher fees for this specialized experience. Its 15% gallery fee is way higher than other marketplaces (X2Y2 only charges 0.5% and aggregators are free). Even its 3% secondary fee is higher than the 2.5% standard set by OpenSea. Anecdotally, it also has extremely high rand loyalty.
SuperRare thus requires a primary transaction to even enable secondary transactions. If the Primary Transaction Thesis holds and most value is created in initial sales, SuperRare benefits. The fact that SuperRare’s primary sale fee is 5x that of its secondary fee is also some evidence that I’m not the only one who believes more in primary sales.
We analyzed the eBay-Amazon rivalry, but so far omitted another e-commerce winner: Shopify.
When you sell goods on Amazon, you tap into a wellspring of existing demand, and fulfillment by Amazon (FBA) helps you hand off logistics. But Amazon’s demand requires a sacrifice: You have to play by Amazon’s rules, pay Amazon’s fees and appease Amazon’s algorithm.
Having your own Shopify store gives you more freedom. Shopify’s customization lets you build a brand, collect emails, etc. and gives you more control with fewer dependencies. But those privileges come with the responsibility of demand generation—no Amazon search traffic is coming for you.
If we predict physical ecommerce trends like the primary transaction shift to emerge in NFT commerce, we should also see a Shopify analogue for NFTs, right?
Vertical marketplaces aren’t new. In fact, the collection-exclusive CryptoPunks marketplace was one of first NFT marketplaces ever. From custom-built marketplaces like Scapes to self-serve tools like Rarible’s community marketplaces, NFT collections seem to want independence from OpenSea’s fees, rankings and brand.
Comparing vertical NFT marketplaces to Shopify stores is salient, but hand-wavy.
It’s easy to claim “Trading on a collection’s own marketplace is like buying a from a brand’s own Shopify store”. Both give brands more control over branding, data and features. But the economics differ:
Vertical NFT marketplaces suffer the same chicken-and-egg liquidity problem as any other secondary marketplace. Instead, Shopify stores sell new goods, aka primary transactions.
Rather than setting up a brand’s Shopify store, building a vertical NFT marketplace is like setting up a brand-exclusive flea market: You rely on owners to list things on your platform. Shopify stores don’t.
That doesn’t nullify the benefits of running your own marketplace. You can import liquidity with aggregation protocols. This makes vertical marketplaces mostly front-ends to other marketplaces, which weakens the Shopify comparison. Shopify helps you pay fewer fees and choose your own prices. Neither is true on vertical NFT marketplaces—you can’t control prices on secondary sales and you depend on a liquidity protocol for fees.
Having your own secondary marketplace is useful, even if you have zero native listings: You can build an on-brand experience, portray the artwork/media in the right context, capture more data, build your email list, etc.
These benefits are valuable and something brands might pay for. But with an incoming shift to primary transactions, capturing secondary trading fees on one/a few collections is unlikely to be the home-run OpenSea is.
The real “Shopify store of Web3” is the minting site.
Minting has rarely happened on marketplaces. With NFT companies and creators having every incentive to focus on primaries, I believe that’s unlikely to change.
While secondary markets will keep existing, the post-PTT NFT ecosystem will look different.
It’s hard to predict a market where use cases, sentiments and valuations morph. Perhaps NFTs join the Beanie Babies, Furbies, Tamagotchis in the ranks of fads we can look back on and snicker at our own financial imprudence.
But let’s assume web3 has a future after The Great Subtraction in which there’s organic demand for NFTs. Which marketplaces will prosper in the post-PTT world?
It’s hard to predict who’ll win. But if a shift to primary transactions is underway, we’re headed for big changes. SuperRare seems great at leveraging the perhaps only undeniable use case for NFTs (art).
But if we go back to the eBay-Amazon rivalry, we shouldn’t look among the obvious contenders. Instead, we need to find the equivalent of the inconspicuous online bookseller about to skyrocket and capitalize on the new demand.
Perhaps it’s the primary-focused minting platforms like Manifold and Bueno that win the future of NFT trading. I’d keep an eye on them.
Perhaps it’s something completely different. And while the casino is closed for reconstruction, I hope you found this essay helpful top understand where the NFT market is headed.
👋🏻And that’s a wrap. I hope you enjoyed this essay and learned something from it. If you want to hear about future essays, please click the subscribe button below, follow me on Twitter.
If you want to amplify this message, share this essay with more people, by retweeting the tweet below:
Finally, if you want to offer some support for independent, long-form and in-depth writing, please connect this essay as an NFT below. I’ll soon launch a Mirror NFT-holder-exclusive private Telegram chat where you can ask me questions, see WIPs and propose ideas.