The superpower of crypto is creating new assets and markets.
As a consequence, one of the most common and successful business models in crypto is the exchange model. If you’ve been in crypto for some time, this isn’t necessarily surprising, but I think it's often under-appreciated when evaluating crypto companies and protocols.
Exchanges can appear in a variety of instances but generally create a venue for transactions for some asset or service, and a mechanism to process those transactions.
Nowadays, I try to ask the question – how could this protocol or business become an exchange? For certain businesses it’s quite clear, but for others, it requires some imagination.
One general rule of thumb is – the closer proximity of the app to the transaction, the more likely that a budding app can blossom into an exchange. This is intuitive for applications that perform transactions themselves (e.g. swaps) but is also true of applications that sit in privilege positions of the stack (e.g. are in the flow of value) or that start as SaaS products but can grow into platforms with robust marketplaces.
Exchanges are well-positioned to develop in scenarios were:
New assets emerge onchain
Apps control distribution and can introduce transactional behavior
New services emerge that impact valuable onchain state or are somehow connected to transactions
Crypto games control their own asset issuance and have open economies
Developer platforms can introduce service marketplaces or auction houses for transactions
When new assets emerge onchain
This is probably the most obvious instance in which an exchange is ripe to emerge. Coinbase created a marketplace for bitcoin. Early entrants like Etherdelta showcased the market demand for trading long tail tokens, although second mover Uniswap emerged as the dominant venue. The list goes on for both newly created assets NFTs (i.e., Superrare, Opensea) and for derivative assets like stablecoins (e.g. Curve) and perps (i.e., Deribit, dydx, etc.)
To be sure, not all exchanges succeed or have similar moats. Some new assets such as ERC 1155s or Bitcoin ordinals have been best served by incumbent exchanges (e.g. Magic Eden) rather than newcomers. Exchanges, based on new assets, are typically most defensible when they have a combination of liquidity network effects, control end user distribution, or are the privileged issuer of the underlying assets that they trade.
Apps that control distribution
Generally speaking, controlling end user distribution enables an application to level up into an exchange. Phantom and MetaMask are wallets, but the advent of MetaMask Swaps turned all wallets into exchanges. Telegram bots are exchanges, but located in a more convenient place for users – their DMs. Social apps like Farcaster, Lens, and Unlonely offer built in transactions for exchanging assets – memecoins, NFTs, points, etc. – at the point of distribution. Similarly, Solana blinks now offer a model for any app to implement in-app transactions and become exchanges.
In the future, I expect more distribution focused apps to offer in-app transactions leveraging primitives similar to Solana blinks, Farcater frames, or Lens open actions. Historically, web2 apps were forced to direct users to transact on other platforms, but now blockchains can offer transactions at the point of distribution, meaning all apps can be marketplaces.
Marketplaces for new services
Service providers that impact valuable onchain state or are somehow connected to transactions, have the ability to become exchanges. For instance, oracles bring external data (e.g. stock pricing data) onchain, which can impact the onchain price of other assets, trigger liquidations, or result in arbitrages that generate instances of oracle extractable value (OEV). Oracle providers like Pyth step into an exchange model by creating auctions that bundle transactions with oracle updates and let people pay for priority in back running transactions. Bridging and cross chain interoperability services naturally act as exchanges, by similarly controlling pricing of assets
More emergent are ZK proof marketplaces and services that must be submitted on-chain. As a result, ZK proof marketplaces (e.g. Gevolut) and aggregators (e.g. Nebra) will have to compete with transactions and other proof generators for scarce blockspace. While the proof service landscape is emerging, the leading providers will benefit from economies of scale (more proofs → more aggregation → cheaper proving – more proofs) which will potentially allow these service marketplaces to become valuable exchanges.
Crypto games
One notable difference between web2 gaming and web3 is the shift towards open economies with transferable assets. Since crypto games like Axie often control asset issuance and velocity (e.g. turnover) they can verticalize their own in-game exchanges or otherwise act as asset brokers. This naturally provides web3 games the opportunity to introduce an exchange business.
Developer platforms as marketplaces or auction houses
Developer platforms are a little less obvious, but benefit from economies of scale and possess privileged positions in the stack that put them in close proximity to transactions.
Rollup as a service providers (RaaS), such as Conduit and Caldera also share the ability to insert themselves into the value flow by creating their own sharded sequencers that become a focal point for transaction ordering for RaaS supported chains. While shared sequencers can certainly order blocks themselves, they can alternatively auction transactions and capture MEV.
Another core piece of developer infrastructure are wallet as a service providers (i.e., Dynamic, Crossmint, Privy) which have the ability to connect user funds across apps providing a sticky user and developer experience. With this lock-in, WaaS providers offer in-app transactions, swaps for on and off ramp, and possibly offer similar wallet swap plug-in features.
As the crypto economy grows and all assets move onchain, we’ll undoubtedly see many more exchanges emerge.
Exchanges are everywhere, some in plain sight, some waiting to be uncovered.
If you’re building a new type of exchange, my DMs are always open.
Thanks to Franklin Bi and Zaid Fattah for feedback on this essay.