May 17, 2022

10 things I’ve learned designing for decentralized finance (DeFi)

  • February 7, 2022
  • 13 min read
10 things I’ve learned designing for decentralized finance (DeFi)

I’ve worked in crypto, on and off, since 2017. Here are a few things I’ve learned along the way.

A timeline showing milestones in the history of cryptocurrency

Image from Finematics:

1. Progress is non-linear and non-uniform

To an outsider, “crypto” is a single monolithic entity. The word suggests an alternative digital currency. Which, to be fair, is true. It’s just that it’s also so much more. NFTs are digital deeds of ownership. Decentralised Finance (DeFi) is transforming finance. DAOs are changing governance. On-chain identity is something we’re still figuring out.

On top of that, there is the ever-growing framework of apps, protocols, chains, and wallets that support this ecosystem.

An infographic for the web3 stack

An infographic for the web3 stack

To think “crypto” is just money is missing the wood for the trees. Newer buzzwords like “web3” and “metaverse” may induce a knee-jerk cringe, but at least they’re closer to what we’re really discussing here.

Once you get inside this world, you see that there are many things to work on, and a surprisingly small group of people spread across all of them.

Which explains why this giant thing mutates and grows at unpredictable speeds in unpredictable directions.

For example, Ethereum had existed for many years before the launch of Uniswap, the first automated market maker. Once Uniswap launched, decentralised trades could take place and one giant piece of the puzzle was solved. It spawned dozens of imitators.

The popularity of DeFi led to enormous gas fees, but then competing chains appeared and a whole new area of the map was opened up, almost overnight.

I find it amazing that, despite the growth of web3, Metamask is still the main wallet and it is still basically terrible. But one day, there will be a revolution in wallet UX, and a huge stride forward in that direction will happen.

Uneven progress is still progress.

2. “We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run”

Often referred to as Amara’s law, this refers to our trouble conceptualizing exponential growth. Connected to the point above, crypto progresses in spurts. Usually there are lots of little steps where not much seems to happen, then giant leaps forward. If you’re new to the space, you probably joined after one of these growth spurts (various inflection points in the adoption of Bitcoin, DeFi summer ’2020, the NFT craze) and think “THIS WILL CHANGE EVERYTHING” and expect the same pace to continue every month.

The thing is, each of these technologies (Bitcoin, DeFi, NFTs) will eventually change everything (and I really mean that), but probably over a longer timeframe than we imagine and in ways none of us can accurately predict.

I think this law also explains the greed we see when number go up: everyone wants to be rich tomorrow, but few can stomach the years of waiting between these occasional giant leaps. Most people miss out on life-changing gains because they simply don’t stick around for long enough.

3. The most talented, interesting, people in tech aren’t working for Google or Facebook anymore, they’re working in web3

Around 2008–15, I’d say Google was the main goal for super smart, ambitious, young people. The products, the people, the free food, the pay!

Well, the appeal has waned and now they’re going into crypto instead. I’ve found this crowd to be an intriguing, talented, bunch of individuals that make other places look boringly homogenous by comparison.

What’s interesting to me, is that this is now happening across all seniority levels. I can easily see why smart young renegades might graduate from college, look at Google or F̶a̶c̶e̶b̶o̶o̶k (sorry, Meta) and go “fuck that”; but we are now seeing a lot of older executives move over too.

It’s good to see them resist the temptation of “resting and vesting” to pick up a new challenge. Maybe it reminds them of the old days when their company was fun and they got to build cool stuff?

And if you’re a talented, forward thinker — why not join us? You’ll get a growth role in a growth industry. The people are more interesting. The challenges are harder (=more rewarding), and if it’s important to you, you can even make more money.

4. …And that’s the main thing they care about

One thing I’ve noticed about people working in crypto, is that they spend a lot of time time working on crypto and a lot of time talking about crypto. To be honest… the obsession is a bit much sometimes. But it means you get to focus on your work and ignore the distractions. Nobody tells you what to do, you just get on with it and be yourself.

This is perhaps one of the reasons why the rebellious individualists that used to go work for FAANG companies are now flocking to crypto.

Every fact of science was once damned. Every invention was considered impossible. Every discovery was a nervous shock to some orthodoxy. Every artistic innovation was denounced as fraud and folly. We would own no more, know no more, and be no more than the first apelike hominids if it were not for the rebellious, the recalcitrant, and the intransigent.

— Robert Anton Wilson

Being completely decentralised means you also get out of your American/European bubble. My own team comprises members in Africa, Europe, Latin America and Russia. It is common for people to use pseudonyms. A few individuals I work with never turn their camera on.

Identity is obfuscated from the start and privacy is by default, not an opt-in.

I find it rewarding to work with different individuals, from all over the world (even if I never actually see them).

5. Complex systems tend to be fractal

A Fractal System is a complex, non-linear, interactive system which has the ability to adapt to a changing environment. Such systems are characterised by the potential for self-organisation, existing in a nonequilibrium environment. FS’s evolve by random mutation, self-organisation, the transformation of their internal models of the environment, and natural selection. Examples include living organisms, the nervous system, the immune system, the economy, corporations, societies, and so on.

— “A brief description of ‘Complex Adaptive and Emergent Systems’ (CAES)” 
by Peter Fryer and Jules Ruis

A fractal pattern is one that contains similar patterns at any level of magnification. You can look at a part of the pattern and still see a near-identical version of the whole.

In designing for DeFi, we tend to come across a lot of the same problems over and over again, at almost every level.

This leads to thoughts of “Well I can’t fix this until I fix that, which depends on this, so I need to solve that first, but really I’d like to fix this whole user flow, but I can’t yet, because we’re stuck into doing it this way, because that’s the way the smart contracts on this level operate…” It’s a sensation that I’m sure a lot of engineers have felt on complex projects.

At this point, I’ve decided to embrace it. Nothing will ever be finished. Instead, DeFi is an asymptote, whereby further progress will get us closer and closer to the line, without ever reaching it.

A diagram showing an asymptote.

A diagram showing an asymptote.

The curved lines tend towards the axes, but will never touch them

On a bigger level — remember, fractals occur at every level — there are also complex probability problems around how people actually use your apps. The game theory is that the higher the apy, the greater the incentive, but only if inflation or sell-offs don’t tank the reward token’s price. Similarly, the higher the total value locked, the more trustworthy you are, but the fewer people in the pool, the higher the apy. Various parameters can be adjusted, such as inflation, lock times, and reward weights. But you are still appealing to a large number of intelligent and adaptive agents who can choose to play any game they choose.

These are hallmarks of a large complex system.

If we go up from the user level, to the project level, we see the teams themselves are playing a similar game called Curve Wars.

The many layers we find in crypto introduce more curious fractals into the system, as each layer both depends on, and mirrors, the other. While writing this article I read the docs for [REDACTED] Cartel, who obviously noticed the same thing:

Like ETH, the Curve ecosystem has layers. Treat the “L1” in this scenario as Curve. The “L2s” built on top of it, scaling Curves influence, are yield aggregators such as Yearn. However, due to the economic incentives tied to the protocol, the layers keep going. For shits and giggles lets assign the L3 moniker to Convex. You’ll understand why by the time you finish reading — we’re climbing up a new branch of the tree. [REDACTED] Cartel might be the L4.


Many players, many games.

6. The future is multi-chain and cross-layer

Ethereum was cofounded by Vitalik Buterin, Charles Hoskinson and Gavin Woods. The three then split up.

Vitalik Buterin stayed with Ethereum.

Charles Hoskinson founded Cardano.

Gavin Woods founded Polkadot.

Ethereum, Cardano and Polkadot each have different approaches.

Solana and Terra are also launching dApps.

Some blockchains are EVM-compatible (Ethereum Virtual Machine), making it very easy for them to spin up a version of existing dApps. Examples include Fantom, Binance Smart Chain, and Avalanche.

Ethereum itself is scaling via second layers, which offer the same security, but without the congestion.

Interesting things are also going on with the Inter Blockchain Communication Protocol and the Cosmos ecosystem, which aims to connect all these chains together.

I have worked on dApps for Ethereum, L2s, Solana, Polkadot, Kusama and Cosmos.

I don’t think there will be a single winner. Instead, it’s more likely that networks will evolve into different niches. And if there is going to be one single winner, that day is a while off.

Sometimes it is frustrating having to learn yet another wallet and another system architecture, but… See point 1.

7. The average DeFi app has a UX maturity level somewhere between “user hostile” and “developer centred”, but certainly nowhere near “user-centred” or “user-driven”

A diagram showing different stages of UX maturity

A diagram showing different stages of UX maturity

Jakob Nielsen — the OG of User Experience Design — says it can take 20 years for a company to reach the final stages of maturity.

I’d say specific parts of the crypto industry are actually doing quite well. Some companies are at least halfway up the steps, but DeFi specifically? Yeah, not so much.

However, this isn’t necessarily a bad thing. At least right now, because…

8. Insanely complicated degen strategies are fun and addictive

Below is an actual message I received from an acquaintance regarding a particular farming strategy:

Buy fantom lock them for 2 or 3 more months on the fWallet (Staking) get around 6% on ur fantom, secure the network. 
And at the same place you can mint sFTM equivalent to the amount you staked. 
Then you should lock those sFTM as collateral to borrow fUSD 
You can get around 33% of your collateral value. 
Use those fUSD to buy wFTM on fSwap 
(All in the same place) 
Once you have your wFTM transfer them to 
Lock then in their vaults they give you good APY also , don’t have data for now 
Then with the deposit receipt they will give you (yvFTM) bring those to Abracadabra lock them as collateral. 
Take a loan but don’t risk too much, at these prices, i will take only 30% of my collateral. 
Then bring those MIM to buy Tarot token on spirit swap 
Then go to and lend your Tarot single sided for 100 to 120%


One of the most popular games right now is “string as many different things together as I can to eke out as much yield as possible, regardless of how risky or time-consuming it is”.

This game produces a lot of dopamine hits, a sense of superiority, and (if done properly), a lot of money.

In other words, it’s the greatest game you’ll ever play.

Developers of DeFi projects have responded with their own game called “how fucking complicated can we make this protocol, so that degens will ape in and call it a new paradigm?”

And thus we have leveraged yield farms, boosted liquidity rewards, bonds, rebase tokens and ponzinomics.

When everyone playing the game is a pro, why even bother with instructions? Who needs intro text when they’re just looking for the next kick? Why make it easy, when the whole point is that it’s hard?

If it wasn’t exciting, nobody would do it. I think this kind of degen madness promotes creativity, keeps people on their toes, and accelerates the whole industry. It’s not for everyone, but neither is croquet.

Many players, many games…

9. Insanely complicated degen strategies are not very newbie friendly

This is the flipside. I think it’s ok to have degens rushing in, throwing bombs, and clearing the space, but the eventual goal has to be an open, decentralised, financial system, available to everyone.

The next growth spurt we see in DeFi is going to be greatly improved UX for liquidity mining and farming rewards.

Some feedback and error messages wouldn’t go amiss.

Automated strategies are getting better.

Auto-compounders make everything easier.

Safety is being prioritised.

There’s still a reluctance to have any explanatory text, but we’re getting there.

A meme showing a bored woman being lectured about crypto by a guy in a nightclub

A meme showing a bored woman being lectured about crypto by a guy in a nightclub

Actual footage of an ape in the wild

10. Perfect is the enemy of good enough

Open-source is sacrosanct, which means that everybody copies everybody and expects to be copied. See something you like? Fork it, make a few tweaks, launch it. Later on, another team will like your tweaks and copy those. We all progress as an industry and the UX gradually improves.

Things are almost expected to break. DeFi is still the Wild West and everyone knows we’re all just experimenting. So let’s be brave and try stuff!

While I’d love for everything to be perfect at launch, I reluctantly acknowledge its better to just get stuff out there and refine as we go.

There is a fantastic essay from the inimitable Gwern called “Bitcoin is worse is better”. It was written in 2011 after Bitcoin hit the heady heights of $11. Reflecting on its wild success, he writes:

“The basic insight of Bitcoin is clever, but clever in an ugly compromising sort of way.”

That philosophy lives on 11 years later, because there are always compromises, whether we’re talking about Bitcoin, or Ethereum, or [insert favourite chain] or DeFi as a whole. But that’s OK, because if people are using it, we can keep improving, keep iterating, and keep shipping.


About Author

Scott Larsen