Sharing some long-form thoughts on protocol business models, how to evaluate protocol financials, and best practices that we can apply from studying the most successful marketplace businesses of the past.
At this point, almost every L1 has pivoted to an appchain-centric scaling roadmap.
While this is a welcome pivot, I’m concerned that we’ll end up speedrunning through lessons already learned in the evolution of Polkadot and Cosmos over the past few years. 🧵
Most protocols are far too focused on tokenomics.
At best, it’s time that could be better spent elsewhere. At worst, it leads to worse products (or ponzi loops).
In this thread, we’ll lay out some thoughts on what sustainable business models looks like in crypto. 🧵
Most rollups have a RaaS-sized hole in their business model.
Some thoughts on the challenge that rollup frameworks face today, and why they should use M&A to solve it. 🧵
Rollup-as-a-Service (“RaaS”) providers are at the center of one of crypto’s most important questions: who is best positioned to capture value in the rollup ecosystem?
The answer to that question, it makes sense to start by examining RaaS economics 🧵
It’s well known that MEV extraction can accelerate network centralization and harm users through strategies such as front-running. This post compares a few of the approaches being taken to mitigate the impact of MEV in the Ethereum and Cosmos ecosystems.
Cosmos needs to 10x its urgency around distribution.
It has a few projects that will be best in class for their sector (e.g. Osmosis, dYdX).
But you can't create the "wow" moments that drive organic adoption without getting these products into the hands of non-Cosmos users.
@rphmeier
I think Cosmos social capital has about 12 months to do something unique and differentiated otherwise we get swallowed by Eth flavored variants of Cosmos originated ideas like roll apps and Eigenlayer.
I feel intense urgency. The hour is late.
MEV is top of mind right now for every chain and application.
Sooner or later, every protocol must land on their own solution.
As
@AntonioMJuliano
and others have observed, appchains have a few big advantages 🧵
Extremely excited to announce an initiative that we've been working on for the past few months!
As detailed in the post, Mesh Security will hugely improve the economic security of Cosmos chains and enable a number of exciting new use cases.
In partnership with
@axelarcore
,
@akashnet_
, and the
@ATOMAccelerator
,
@OsmosisGrants
is thrilled to announce an initiative that will fund the development of Mesh Security and bring it to mainnet across the Interchain!
More details below 🧵
The thread below points to a larger strategic question for
@osmosiszone
. It’s ineffective and expensive to do both, so Osmosis needs to decide if it wants to be:
A. The only DEX that can match the asset coverage and feature set of a CEX
B. The primary liquidity venue for Cosmos
Excited to co-host this season of Bell Curve with
@MikeIppolito_
!
We'll explore all flavors of app-chains and their various trade-offs with the help of some of the leading builders in the space. Tune in next week for our chat with
@sunnya97
and
@dberenzon
!
The season 3 opener of bell curve on the appchain thesis is live!
This season my co-host is
@MylesOneil
, one of my good friends but also one of the smartest people on the subject I know.
Enjoy!
Apple:🎙
Spotify:🎙
Ok crap – we have two issues:
1) Recruiting high quality sets of sequencer operators and bootstrapping collateral to secure them is actually really expensive.
2) These apps want atomic composability or at least guaranteed inclusion for cross-chain txs.
To be more specific, I see an increasingly common disconnect between the capabilities being marketed, the solutions proposed to deliver on these promises, and the realistic trade-offs they'll have to make based on what we've seen from solutions already in production.
I’ll stop there because this thread is long enough but I hope you see the point that I’m trying to get across.
Almost everyone now agrees that in a perfect world, you could offer the sovereignty of a Cosmos appchain with the security properties of an L1 like Ethereum.
@hasufl
Took a shot at mapping these to P&Ls (Maker is understated until MKR expenses are added)
Lido's GM % looks like a healthy SaaS business, Maker looks like financial services
This changes if you consider the 1.5M LDO incentives to Curve to be COGS (liquidity maintenance cost)
For the nerds still here, I’d encourage you to check out our podcast episode with
@KAndrewHuang
from
@conduitxyz
.
We dig into the value prop of RaaS, the competitive dynamics between different stakeholders, and a whole lot more.
Let’s start with a simple observation: token appreciation is not a business model.
Financial engineering can be useful in particular cases (e.g. EIP1559 burn mechanism), but only as a complementary feature of an economic system.
We like Polkadot’s shared security in theory, but it's likely unrealistic to expect an L1 like Ethereum to build this into the protocol.
And we want the benefits of Cosmos sovereignty and IBC, but there are a few things we don’t like about Cosmos and want to avoid.
Ah crap, now we have a cost problem.
State-aware sequencers are a lot more expensive so it doesn’t seem like state-aware shared sequencers can scale beyond ~10-20 chains.
Wait, doesn’t this sound like the issues that the Cosmos Hub is facing with its consumer chains?
This is absolutely what we should be striving for, but it feels like many ecosystems are underappreciating the trade-offs associated with appchains and think it will be easy to “have their cake and eat it too.”
But if it turns out that we haven’t yet discovered the silver bullet to scaling, I hope we can save ourselves some pain by looking at the past with a better appreciation for the design choices and associated trade-offs made by first-gen appchain builders.
Here is how the internal monologue seems to go for many projects:
Gen purpose chains don’t scale and appchains enable better products.
Ok, appchains are the way!
We don’t want our chains and interop channels to be secured by their native token.
And unlike Cosmos, we want our L1 to play an important role and accrue significant value to the token so it can continue serving as the schelling point for alignment and coordination.
Ok we got it!
We’ll have multiple base layers (e.g. OP mainnet, Starknet) serve as messaging hubs to enable composability and we’ll have appchains secure themselves using the base layer's token, which should drive demand and accrue value through fees / rev share.
And to be fair, there are definitely projects that are not cutting corners in their marketing and working to develop novel solutions to these problems (e.g.
@sovereign_labs
,
@anoma
,
@ArgusLabs_
).
Now we arrive at the current state of the world. Sound familiar?
Most of the ideas I’ve heard sound a lot like lightweight versions of the Polkadot security model (e.g. restaking) combined with the elements of Cosmos hub and spoke approach.
They may look at Polkadot and think they can replicate the benefits of shared security without the pain that parachains have to deal with.
They may look at Cosmos and think they can replicate the benefits of sovereignty and IBC without the scary security tradeoffs.
@conduitxyz
started by aligning themselves to a single framework in the OP stack.
By going deep instead of wide on framework support, they were able to get a boost to distribution from OP as well as being able to focus all resources on building one, truly great product.
But my sense is that the solutions being marketed are either a) not as scalable as folks think they will be or b) present a direct tension with the benefits of sovereignty that you get from appchains (which led to these pivots in the first place).
📣 Announcing the EigenLayer Research Fellowship (ERF)
The EigenLayer Research Team currently has bandwidth to work with 5 fellows deeply on a topic that they care about and interests us 🤝
I’m not saying that every design choice made by “first-generation” appchains was correct or could not be improved upon.
There are definitely second-mover advantages that these projects should try to capitalize on.
Wait, we have a state-awareness problem:
These appchains want their sequencers to be state-aware so they can optimize blocks and do fancy MEV things.
And the shared sequencer needs to be state-aware to guarantee the validity of tx’s which we need to guarantee inclusion.
Here’s a solution:
We’ll have different chunks of the base layer val set handle the standard tx’s.
We’ll get 3rd party shared sequencer networks to handle the cross-chain txs.
And both can be done cheaply by limiting the role to ordering tx’s as they come in!
While I’m firmly in the camp that that issuance = cost for networks, over-cutting on costs could lead to outcomes that undermine the value prop of the network.
@sachayve
does a great job of exploring this trade off space in his latest piece. Highly recommend a read!
some thoughts on mvi
tl;dr: while minimum viable issuance is, by definition, a good idea, if we are to minimize the risk of unexpected negative externalities we need to start viewing viability through a more multidimensional lens.
Of course – I could be very wrong.
Maybe centralized sequencers are the end state, sovereign rollups can replicate all the capabilities of a full-stack appchains, and coordination is improved through the alignment of building on a shared L1.
@luigidemeo
Seems like many are on step 2
1. Ok, appchains are the way
2. Shared seqs will get us composability
3. Wait, we need state-aware seq's to optimize blocks and do fancy MEV stuff
4. State-aware shared sequencers can get us both
5. Wait, this has the same scaling issues as ICS v1
Really enjoyed chatting with
@KAndrewHuang
and
@bcmakes
about the rapidly evolving RaaS space.
Similar to AWS's role in Web2, RaaS providers have a huge opportunity to expand the market of rollup devs by offering production ready components that can be easily pasted together...
The newest episode of
@ReveriePod
features
@KAndrewHuang
, the founder of
@conduitxyz
, as well as
@bcmakes
.
We explore what Conduit is building, RaaS dynamics as a whole, and dive into what the future of the rollup space might look like.
@ttunguz
did some quick back of the napkin math to help answer your question. you're right in that the numbers don't look good vs b2c SaaS BM's, but there's some important considerations to keep in mind:
@MylesOneil
@realWilliamFu
if that math is correct, then a customer who is referred and unstakes before the 18 month mark is contribution margin negative, no?
Before diving deeper, I’d encourage you to check out our long-form thoughts if this topic is up your alley.
For everyone else, I’ll do my best to provide the tl’dr version below.
Perhaps because crypto lacks these frameworks, few teams track their business models and associated financials closely.
That often leads to poor decisions and business practices.
As Bezos says, even if it’s a 10% chance of a 100x payoff, you should take that bet every time.
It may be the harder path, but I strongly believe Osmosis should take a bet on itself and try to become the largest version of what it can be.
New episode of Bell Curve is live!
We took a deep look at the Cosmos approach to MEV, including:
- MEV value chain
- ETH vs Cosmos approach to MEV
- Separation of OF vs blockspace mkts
- What makes “bad MEV?”
Enjoy 🙏🏼
Apple:
Spotify:…
I see a large opportunity for infra and service providers developed with app-chains/rollups in mind from day 1
Incumbents can’t charge the L1 level pricing required support their cost structure, and they aren’t nimble or motivated enough to rebuild their architecture
EVM is the de-facto industry standard when we are talking about Smart Contract platforms, which in terms means that Ethereum should continue to be the leader among Web3 app developers.
This also means it will continue to have the most value locked compared to other Smart…
Really enjoyed chatting with
@sunnya97
and
@dberenzon
!
We dive deep into the motivations and trade-offs of appchains, considerations around blockspace permissioning, the future of interoperability and much more!
But Osmosis has 2 advantages over both Uni and CEX’s:
- Unlike Uni, Osmosis can offer coverage for almost every major asset as well as an integrated margin engine via
@mars_protocol
in a single UI.
- Unlike CEX’s, the protocol is transparent and controlled by its community.
We're live! Incredibly excited to support
@OsmosisGrants
with the
@hi_Reverie
team. The OGP represents an open invitation for any teams or individuals that want to help accelerate the growth of Osmosis and its ecosystem. DM's are open for anyone that wants to get involved!
Something to watch closely — can OP Superchain add support for interoperability/sovereignty faster than Cosmos can build brand/distribution/ecosystem?
Generally, my sense is brand/distribution/ecosystem are significantly harder to reproduce than good technology/ideas.
@MattFiebach
Removing the absurd bridging fees will be a big step.
But that still assumes the user is 1) already crypto native, 2) knows about Cosmos and 3) wants to use a Cosmos app
This is still too small of a pie. Acquiring new users before they even try Ethereum should be the goal.
Excluding applications, it looks like 3 major stakeholders will compete to capture value in rollups:
1. Rollup frameworks
2. RaaS + infra services
3. Shared sequencers
Maximizing marketshare:
Bootstrapping liquidity for long tail assets has been a great way to break into the market for Osmosis as it was for
@Uniswap
.
But over time, Uni shifted its focus to large caps + stables which has allowed them to compete with CEX’s in volume.
Studying the economic attributes of the most successful traditional marketplaces businesses and thinking about how to apply those lessons to your protocol is a great place to start.
I’d argue that the majority of protocols most closely resemble software marketplace businesses.
Similar to marketplaces, many protocols provide a service by connecting the supply side to the demand side in order to create new markets.
@0xShitTrader
I think this is a big challenge for gen purpose chains given their value prop.
The obvious answer is via permissioning at the L1 layer like we see with Cosmos appchains.
But for ETH, the cons of intro'ing formal governance to express these opinions far outweigh the benefits
With applications in particular, these mechanisms typically lead to short-term, narrative-driven price pumps followed by a return to the mean.
In the long run, these strategies are not a substitute for a sustainable business model.
There are plenty of low hanging fruit improvements to be made in crypto.
I suspect that in the years to come, a big differentiator between the winners and losers will be those with a firm grasp of their business model vs those that don’t.
A good place to start is to think about the economic qualities of a protocol – the product, customer, cost structure, and revenue model.
With this context, you can then look at existing companies with similar business models to see how they work out in practice.
@sunnya97
@aidan0x
Implicit permissioning via a heavy tax tx’s for non whitelisted contracts could be a happy middle ground
You don’t want NFT mints clogging up a DeFi chains blockspace as we saw with Solana / Serum
What drew me initially to crypto was that DeFi and blockchain technology could actually deliver on fintech's original promise to reimagine traditional finance.
Fintech simply put a new sheen on old rails. Crypto represents entirely new rails. It will take time, but the world…
@hxrts
@MikeIppolito_
the one that everyone owns and doesn't fluctuate in value* sounds like the logical choice to me :)
(*excluding the occasional bank run crisis)
@AlanaDLevin
It’s a choice just like appchains (e.g. Optimism’s whitelist), tho would note that we haven’t seen it curated through governance yet
Also will be interesting to see at what point permissioned appchain deployments actually get contentious
The question then becomes, what types of crypto business models exist and how can you assess their quality?
We go into far more detail in this piece, but I’ll try to summarize our thinking as best I can in the rest of this thread:
Sharing some long-form thoughts on protocol business models, how to evaluate protocol financials, and best practices that we can apply from studying the most successful marketplace businesses of the past.
@mattimost
@hasufl
I think its a mix of COGS and CAC (OPEX) spend today
COGS: cost to ensure sufficient liquidity on DEXs (Curve) for users and integrations (Aave). How fast this trends to zero after withdrawals are enabled is important
CAC: incentives on L2s or other L1s, referral program, etc
These stakeholders will fight over two sources of revenue that represent the lion’s share of value capture:
1. Direct execution fees (sequencer)
2. Production ready, hosted infra services
@Flowslikeosmo
@osmosiszone
Wouldn’t it be great if someone could measure the ROI of these investments to understand where Osmosis is overpaying and where it is underspending for liquidity?
A few that come to mind on the Cosmos side are
@scanworks_
,
@SkipProtocol
,
@Fair_Block
,
@NumiaData
But very interested to see how this develops on the Ethereum side.
If you’re building infra and services to support this appchain future, we’d love to chat!
Uniswap is a financial marketplace that connects traders (demand) with liquidity providers (supply).
AirBnB is a physical marketplace that connects renters (demand) with property owners (supply).
@Calderaxyz
has been framework agnostic, aiming to offer customers with max optionality at onboarding.
By operating rollups on multiple frameworks, they can see which components are most valuable and incorporate these learnings into the guidance they provide to new customers.
@nosleepjon
@Hyperlane_xyz
Very interested in Hyperlane!
Seems like it falls somewhere in btwn the shared standard approach (eg IBC) and the solution on top of many standards approach (eg Axelar).
Is that a fair read?
@MonetSupply
@EffortCapital
@RoboMcGobo
Aren't there are some far bigger issues here?
1) This revenue is needed to support the $25M/yr DAO expense (they don't have a "17.5 yr runway" as OP suggests)
2) even when they become profitable, there are cap allocation options w far higher ROI to maximize LT token value
For example,
@osmosiszone
is implementing a number of systems that aim to 1) eliminate harmful MEV extraction entirely, and 2) use on-chain protocol controlled mechanisms that allow the protocol to capture the value of benign MEV extraction.
@ttunguz
did some quick back of the napkin math to help answer your question. you're right in that the numbers don't look good vs b2c SaaS BM's, but there's some important considerations to keep in mind:
As we’ve already begun to see, it seems likely that the majority of startups will go directly to RaaS providers to outsource the complexity of operating a rollup.
If this trend persists, RaaS providers will own the top of funnel as well as the customer relationship.
@EffortCapital
TBH I don't understand this attitude.
If teams build on the Cosmos stack bc 1) it enables the best version of their product and 2) they don't need to deal with the technical or economic constraints of building a rollup then Cosmos should absolutely be embracing this approach.
Maximizing Cosmos adoption:
The best way for Osmosis to increase Cosmos mindshare is to demonstrate that you can build the best version of your application as a full-stack appchain.
And the best way to demonstrate this claim is to capture the marketshare required to back it up.
Using this framework, you could start to assess a business like
@Uniswap
similarly to how you’d assess AirBnB.
It’s not exactly the same, of course, but the marketplace model roughly maps over to most protocols:
With these qualities, Osmosis can offer the closest thing to a CEX experience while retaining all the benefits of a DEX.
In light of everything we’ve seen in the past cycle (e.g. the rise and fall of FTX), that’s an extremely compelling value prop!
@AlanaDLevin
One advantage that the Superchain could have is around coordination of adopting / upgrading standards (e.g. AA, interop, sequencing).
All of Cosmos uses the SDK and IBC ofc, but getting everyone to coordinate on version upgrades + implementation standards has been a challenge.
That said, there are very big questions about value accrual and potential competition with the underlying stacks.
We dig into both the value prop and these competitive dynamics in this episode.
Highly recommend a listen to anyone building in the rollup space!
This might be the minority view right now, but I think betting big on option A is the path to maximize 1) Osmosis’s long-term potential marketshare and 2) adoption of the Cosmos stack.
Hot take:
I think
@osmosiszone
should stop incentivizing pools for major bridged tokens like FTM, BNB, ARB etc.
Yes, those incentivizes result in pools deep enough to facilitate trading on Osmosis,
No, I do not think that it is worth it.
Here is why🧵 👇
#FlipsideAmbassador
Despite this, the current conditions are extremely favorable for rollup frameworks to consider M&A or a large investment in a RaaS provider.
In fact, I don’t think they will ever be as favorable as they are today for rollup frameworks.