First direct, then emergent
|Tony Sheng||Feb 25, 2019|
Last week, I opened up a MakerDAO1 CDP for the first time. By depositing ETH as collateral, I received a loan of MakerDAO’s stablecoin Dai. It was easy2. I don’t think I’ve ever said that about a dapp3.
But more importantly, it was useful. A CDP gives me instant leverage on my ETH. I now had 66%4 more working capital (my original ETH holdings, now locked up in the CDP, and the Dai I borrowed). Time spent? Five minutes.
Now imagine alternative ways to get leverage. How long would it take to get a loan from a bank? From a friend? Or even from a crypto exchange like BitMex? Buying a financial product that gives you exposure to a levered position is probably the fastest way to get leverage in the legacy financial system, but even that would take longer than opening up a CDP.
Easy and useful.
Outside of crypto, it’s assumed that a new product or service has to be more easy or more useful (ideally both) than competitors to have a chance. The challenger must offer something ten times better to stand a chance5. Think Uber vs cabs, Dropbox vs external hard drives, Slack vs email.
The properties that make a user say something is “better” are direct properties. The user can directly observe them. There’s instantaneous feedback. With Uber, you go from waving your hand on the street to tapping your phone and having someone pick you up.
In crypto, many projects and investors bet that users would prefer products for their emergent properties–properties that emerge out of complex systems, often subtle and hard to discern. With Uber, you go from cab drivers not knowing who you are to Uber having a vast database of your locations. Because of Uber’s underlying system, the emergent property of privacy loss comes with the direct properties of convenience and cost.
Users would, the theory went, accept some inefficiencies (e.g. cost, speed) for properties like censorship resistance, transparency, and interoperability. And some users did, just not many. As I wrote in “Growth is not price”, users that value emergent properties can be labeled “ethos users” and they are in the vast minority:
Ethos users (true believers) vs ethos-agnostic users is a useful way to think about addressable markets for crypto projects. To obtain growth, a project can either focus on serving ethos users or ethos-agnostic users.
Ethos-agnostic users are hard to serve. They don’t care about hypothetical benefits. They care about real benefits and direct solutions.
In other words, ten times better on emergent properties isn’t sufficient to compete for ethos-agnostic users. One should aim for ten times better on direct properties.
Ten times better on direct properties is hard to achieve, especially for something built on a blockchain. But if a product or service built on blockchains with beneficial emergent properties can compete on direct properties, it gets stronger over time. A user comes for the direct benefits (e.g. get leverage) and stays for emergent benefits (e.g. permissionless, privacy preserving, censorship resistant).
And by the way, this isn’t limited to MakerDAO and the Ethereum ecosystem. I see this happening with, for instance, Bitcoin as well. Lightning network apps (e.g. lightning pizza, tippin.me) are up and running, competing on direct properties, built on a foundation of the strongest censorship resistance in the world.
As I wrote to members last week:
a powerful one-two punch: come for the direct properties, stay for the emergent ones. […] To grow a permissionless network, hook users with direct properties and convert them with emergent ones. To widen the funnel and attract more users, offer more direct properties (like a savings rate on Dai). To spawn true believers, build systems that are open and permissionless and unable to marginalize any groups.
When we look back at this period, I predict we’ll see a pattern of user growth catalyzed by strong direct properties and user retention afforded by strong emergent properties. We need both, just at different times.
Disclosure: I have no MKR holdings and nominal Dai holdings. ↩
With the exception of Decentraland dapps (e.g. LAND marketplace, Agora) which I think are great but I am biased so don’t trust, verify. ↩
This is at a collateralization rate of 150%, but as I mentioned before, I got liquidated instantly, at this rate, so you should collateralize at 250%+ probably, which would give you ↩