Why The Winners Won

Introduction

With more performant infrastructure on the horizon (rollups, faster L1’s/L2’s, account abstraction), we expect to see a lot more apps launching. That’s why lately, we’ve been thinking a lot about apps. More specifically, why did the last cohort of apps like Uniswap, dYdX, and Lido win while competing apps lost? 

It’s always hard to pinpoint exactly why the winners won, so we’d never say what we’re about to do is a scientific process. Rather, it’s our personal exploration of why these apps won what was an incredibly competitive battle for users — one that has many lessons for future application teams.    

Uniswap

Uniswap is a decentralized exchange (DEX) founded by Hayden Adams in November 2018. With over $5B in liquidity, $1B in daily volumes, and over 100k daily active users, there's little doubt Uniswap is the leading spot DEX in the market. 

Often, you hear people say that Uniswap won because they had the best technology/mechanism design. In our view, the tech played a role, but it wasn't really unique to Uniswap. In fact, the idea for automated market makers (AMMs) — the invention that made Uniswap work so well — was borrowed from a post by Vitalik (and first conceived by Alan Lu from Gnosis Safe). 

To expand on this point further, there were many competing AMM’s, many of which had far more sophisticated and flexible designs than Uniswap. Yet they attracted few users and mindshare (sorry, but researchers on Twitter don’t count).

So if it's not for the tech, how did Uniswap become the winning DEX?

Design and brand

When Uniswap launched, EtherDelta was one of the only DEX’s with real traction. The main problem with EtherDelta was that it was complicated to use, especially for retail users who weren’t used to trading with an order book. 

In contrast, Uniswap was simple to use from the get-go — both for the users taking and supplying liquidity. People who wanted to buy tokens didn’t have to think about limit prices or order books, and people who wanted to supply tokens could deposit them with one-click, which led to a great selection of assets available on Uniswap. 

Moreover, Uniswap had that memorable brand we all know and love: the pink unicorn is unforgettable, whimsical, and aligned with crypto’s frisky culture. One recurring lesson the market teaches is there is great value in building a recognizable brand when you’re selling an easy-to-replicate product to retail users (a great brand instantly differentiates you from competing products offered by competitors).

Those who were around during the early days of Uniswap will remember how viciously professional traders attacked Uniswap’s design for not being efficient compared to traditional order books. But as user behavior has shown, retail users do not choose exchanges based on efficiency/best price/feature set. Rather, they care about how simple the product is to use, brand, and asset selection — things Uniswap got right at launch.  

Amazingly, to this day, Uniswap's core interface hasn't changed much from the demo released in 2018. Great UX patterns are rare, and they can be very sticky. 

Distribution

Another observation we’ve had about Uniswap is that an amazing product isn’t enough — you also need to get people to find and use it. 

This point is often forgotten, and one worth highlighting: Hayden hustled to expand his crypto network and get Uniswap in front of the right people early on. He went to conferences and co-worked with known teams like Balance and Maker to demo the product in person. This hustle helped him put Uniswap on the map. In fact, it was during this time that Hayden met Vitalik, who was the person that gave Uniswap its catchy name (in addition to becoming a crucial early promoter for Uniswap). 

On top of demoing Uniswap, Hayden also asked people in his network to help bootstrap liquidity. Thanks to that, Uniswap received around $30k in deposits on the first day, which was just enough for people to start making small ~$100 swaps. 

With the supply-side bootstrapped, early supporters started to create some hype when they realized providing liquidity could be profitable thanks to the fees generated from constant arbitrage opportunities, which led to more people providing liquidity. Kyber's Uniswap integration was another big early catalyst that immediately provided a huge bump to volume, which grew liquidity further. 

dYdX

Founded by Antonio Juliano in July 2017, dYdX is a protocol for trading perpetuals (perpetuals are a type of futures contract that doesn’t have an expiry date). With over $2B in daily volume, dYdX is arguably the most successful decentralized derivatives exchange in the market. 

An often cited view seems to be that dYdX managed to attract users and large volumes exclusively through token incentives. While that was no doubt a large part of it, we don't think they'd be where they are today without nailing two other important things. 

Being responsive to user demands

To build a truly great product experience, the dYdX team regularly adjusted the products they put out to what users wanted.

The history here is interesting. The first product dYdX shipped was Expo, a trading app that made margin trading as simple as possible. By tokenizing short/long margin positions on ETH, all users had to do to lever up was to buy these leveraged tokens instead of having to learn all the complexities of centralized exchanges such as Bitfinex. 

Unfortunately, Expo wasn’t a big hit. The dYdX team realized they had a sophisticated user base who wanted an experience that resembled that of a full on centralized exchange. So they listened to their users and launched a new product with a similar look and feel to that of a centralized exchange. The new product showed immediate organic growth, increasing daily volumes by 20x. 

They were equally quick to scrap the lending, borrowing, and margin trading protocol they had built to focus on perpetuals trading when they saw the tremendous success Bitmex — one of their centralized competitors — was having with them. 

In both cases, dYdX wasn’t afraid to kill a product that wasn’t working to make one that users did want.

Vertical integration

Another interesting thing the dYdX team did was vertically integrate its stack to remove scalability bottlenecks and gain control over crucial aspects of their product.

Early on, dYdX hosted its order book on 0x Protocol, which led to issues such as low liquidity and high failure rates for trades. To solve this, dYdX built its own off-chain matching engine to own their liquidity and reduce failed trades. 

Later on when defi summer caused a spike in gas fees on Ethereum L1, dYdX migrated to their own L2 on Starkware so that traders could enjoy essentially zero gas fees. And, importantly, the increased throughput from this move also allowed them to switch from isolated margin to a cross margining system, which boosted exchange liquidity and made it cheaper to launch new markets. 

And most recently, dYdX vertically integrated its full stack by migrating to their own Cosmos appchain. The move gave them control over their own validator set, which was necessary to build a decentralized offchain orderbook and matching engine. Having a sovereign appchain also gave more power to dYdX community governance, as they can now do things like roll back security incidents and penalize bad behavior such as MEV extraction to protect users.

Vertically integrating more of their stack over time has given dYdX greater control and flexibility to improve UX and product quality. 

Lido 

Lido is a liquid staking protocol founded in 2020 by Konstantin Lomashuk, Vasiliy Shapovalov and Jordan Fish. The protocol's liquid staking token, stETH, allows users to earn staking yields without sacrificing liquidity. With $29B in deposits, Lido is by far the largest defi protocol by TVL today.

Many believe that Lido won the adoption game because of their first-mover advantage. It's true — they were early. But they weren't the only liquid staking token around when Ethereum first enabled staking. In fact, they weren't even the only solution with "stETH" as their ticker. 

Beyond being an early mover, there are several other reasons why Lido has been so successful. 

Bias for action 

Lido's bias for action led them to make certain tradeoffs that allowed them to launch something simple early on and go-to-market faster than the competition. 

When Lido launched, 0x01 withdrawal credentials weren't yet available on Ethereum, which in plain English meant that validators couldn’t select a smart contract as a withdrawal address. This was an issue because smart contracts were needed to create a multi-sig that managed assets securely — a prerequisite for building a product like Lido. To avoid being blocked by this, Lido came up with a temporary solution: they used a distributed threshold encryption scheme to generate a withdrawal address that functioned like a multi-sig. 

Another good example of Lido’s bias for action is their decision to launch with a permissioned node operator design, which involved selecting a group of trusted node operators and giving them the ability to spin up new validators to meet stETH demand without risking any stake of their own. 

In comparison, projects like Rocket Pool went with a permissionless design to maximize decentralization, which required node operators to stake their own capital, thereby limiting the amount of ETH that could be deposited. This created a huge queue to mint Rocket Pool's liquid staking token (rETH), so most users went to Lido instead to start earning yield right away.

Partnerships and integrations

Lido was also incredibly proactive in establishing key partnerships and integrations to bootstrap utility and liquidity. 

Early on, Lido's stETH was way easier to discover and acquire thanks to integrations with popular wallets like Argent and trusted brands like Ledger which helped Lido’s distribution and brand. 

Lido also secured a number of integrations with blue chip defi protocols that further fueled its growth. For example, Lido increased stETH liquidity by incentivizing the stETH/ETH pair on Curve so users could swap between the two with low slippage, which was particularly important as staking withdrawals hadn't yet been enabled on Ethereum at the time. And because of its deep liquidity, stETH was the first liquid staking token to be added as collateral on both Maker and Aave, which increased its utility. 

Conclusion

Revisiting the early days of Uniswap, dYdX, and Lido gives us a good idea of what it takes to build a successful user-facing onchain app. Importantly, it's not just about the tech or being first to market. It's about understanding your users, creating a great product, building trust, and making smart moves. These teams have nailed it (so far), and their stories are a great reference for anyone looking to build a category-leading onchain app. 

Many thanks to Felix Lutsch, Lucas Wotton, and Carl Bergman for their feedback on this piece.