Optimizing Liquidity with Whirlpools

Grace 'Ori' Kwan (Orca Co-Founder) joins The Zeitgeist to discuss design at Orca and how she approaches building for the Web3 ecosystem. 

Show Notes

00:05 - Intro
00:39 - Ori’s background / working at IDEO
04:20 - Human centered design in Crypto
06:03 - The design process at Orca
07:37 - Sourcing for UX research
09:11 - The culture at Orca
10:28 - Qualitative vs quantitative feedback
11:55 - Orca’s success
12:55 - The magic bar
14:02 - Branding
14:26 - Whirlpools 
17:31 - Who is the target user?
19:07 - Branding in Crypto
20:50 - Advice for designers looking to crack into Web 3.0
22:01 - Building a new product at Orca
23:10 - Climate impact work at Orca
24:47 - Working with co-founder, Yutaro

Transcript

Jisi (00:06):

Hi, everyone. Welcome to The Zeitgeist, the show where we highlight the founders, developers and designers who are pushing the Web 3.0 Industry forward. I'm Jisi Guo, design lead at Phantom, and I'm incredibly excited to introduce Grace Ori Kwan. Ori is the co-founder of Orca, the most user-friendly decentralized exchange on Solana. Ori, thanks so much for being with us today.

Ori (00:27):

Really a pleasure.

Jisi (00:28):

Yeah, you are actually in Japan. Am I correct?

Ori (00:32):

Yeah, I'm currently in Japan. I usually bounce around a lot, but lucky enough to be here for the spring.

Jisi (00:37):

Nice. That's amazing. Tell us a little bit about your background. How did you get exposed to crypto and Web 3.0?

Ori (00:45):

It's a funny story, actually. I was never really into crypto for most of my life. In fact, I was one of those people who really didn't care about money in any form besides just using it to survive. But it all really happened during COVID where actually in Tokyo I moved into a shared house, which is pretty common in Tokyo actually, where adults who are looking for a more communal living place will move into this type of shared apartment situation. Right across the hall from me was my co-founder, Yutaro. One thing led to another with COVID and the lockdown. We were all bored, and we were like, "Should we try building something together?" Yutaro was like, "Well, if we're going to build something, I'm only interested if it's crypto," and so it went from there.

Jisi (01:33):

Wow. That's insane. That's awesome. How did you start, were there different side projects that you started with, or was Orca the first project that you thought of?

Ori (01:45):

It definitely was not. We went in with zero intention of ever starting a company. I think because my background's from Stanford, everyone's like, "Oh, you probably wanted to be an entrepreneur your whole life," but that's totally not the case. I didn't really have that plan. I'd always seen an entrepreneurial lifestyle as very tiring and probably something that I wouldn't be suited for. So it really was just like, "Hey, let's start hacking on some projects on the side." Our first project, which we sometimes joke about was really just a little weekend thing. We made it like a little quiz that told you which crypto you were. It was supposed to be one of those Facebook shareable ones.

Ori (02:20):

Then it led to hacking on apps and Ethereum. Then, when defi summer came and the gas fees became really insane on Ethereum, that's when we started really taking notice of the other options available and landed on Solana.

Jisi (02:34):

Nice. That's amazing. What crypto were you on that quiz? I'm just curious.

Ori (02:41):

I think I was Bitcoin. Yeah. Maybe I was more traditionally minded at the time.

Jisi (02:47):

You were the OG. Nice.

Ori (02:49):

Yeah.

Jisi (02:51):

Yeah. Well, it's pretty nice to chat with you because a lot of people know you as a design founder within crypto, and it seems like there are very few designers within Web 3.0 and you previously even have experience working at IDEO. Could you tell me a little bit about that?

Ori (03:10):

For sure, yeah. IDEO was actually a culture shock for me, because I had spent my entire career before that more in engineering-driven places, and I started my career as an engineer. I studied computer science, worked at Coursera, really building fronted applications on the web. Then it was in New York that I started working across design and engineering, this feeling like I wanted to be kind of higher up in the decision-making process, higher up in ensuring that the products we were developing were really going to be useful to people. But at IDEO, it's not entirely a consulting business, but I was working on the consulting side and especially in IDEO Tokyo, that particular office, a lot of it was really about inspiring clients, which is an art in itself.

Ori (03:56):

I learned so much about storytelling, about working with different types of clients and people who have very different mindsets and backgrounds, but it was not as much about shipping products that are really used by people every single day. I quickly learned that my passion is creating products that real people are getting real value out of, which led me to crypto, because nobody is here in crypto just to play around or be inspired necessarily. People want results is my sense of the space.

Jisi (04:30):

Yeah, for sure. I recently read your op-ed in CoinDesk and it was really interesting. You were talking a lot about bringing human-centered design to crypto. Can you talk a little bit more about that?

Ori (04:50):

Yeah. Thank you so much for reading the op-ed. It was actually my first time writing one for this type of magazine, so it was really a fun experience, fun and stressful. But human-centered design, it's a fuzzy thing, but really what I think of it as is focusing on building products that solve needs for people, which sounds so obvious. But when you look at many technology products, especially crypto products, you can really see how the design is accommodating the technology.

Ori (05:20):

As an engineer, it's very easy for me to see how "Oh, well the data structure or the data is structured in this way and it's being returned from the API in this way, so let's just build a UI that spits that out with as little effort as possible to transform that into something that humans can actually make sense of, which results in people just thinking crypto is scary and complicated." At the end of the day, what I think human-centered design is about is keeping in touch with humans throughout the entire process, starting your design and product building journey by thinking about something humans need, continually showing your wire frames, your high-fi designs to people, getting their feedback and thinking very deeply about what the type of person you're trying to build for is.

Jisi (06:04):

Yeah, that's amazing. That's especially relevant because Orca as a DEX, it's a complicated subject matter, but you're known as one of the most easy-to-use DEXs on Solana and being able to really hone in on the user and make that more approachable is super relevant. Could you talk a little bit about the design process that you use at Orca and how do you get insights into what users are trying to do and to make something that they want?

Ori (06:36):

It all starts with interviews. That sounds very formal, but a lot of the time it's really just like my co-worker's friend who happens to really fit the profile of who we're trying to target, and sometimes I like to categorize it this way. For example, we have this friend of a friend like Humphrey, who's a little bit of a Degen, but maybe it doesn't have that many hours a day to spend. So sometimes I'd be like, "Oh, we're designing for a Humphrey right now," or, "We're designing for an Elise right now." So I'll start by like finding the type of people who we think we are trying to cater to and just having a conversation with them.

Ori (07:08):

I think the art of design research is something that I've been very blessed to have experience by working at small startups where designers wear every single hat. So it is about asking these open-ended questions to really just understand what really makes people tick? What they're really looking for, and you won't necessarily get that by just showing someone a design. I think starting actually without even a piece of visual to come from is usually where I begin, so it becomes very exploratory, but then it's about distilling the insights from that open-ended conversation, understanding the needs. Then I go into wire framing, talking to people again, getting feedback, and rinse and repeat all the way until we have hi-fi designs that we deem good enough; not perfect, but good enough.

Jisi (07:54):

Yeah. It seems like you have identified people that are your target personas and you're able to interview them and maybe even observe them and give feedback directly from them, which is amazing. It feels like it's difficult to do that kind of UX research within crypto because a lot of times users are anonymous or you're not quite sure what cohort they may be in. Are there any tips or tricks that you've had and enable to sourcing for that UX research?

Ori (08:25):

I don't hold our candidate sourcing to the same level of fidelity that a very large firm might, because I think it's honestly not necessary. A lot of the types of people that we're looking for are often in our circles, just because maybe I wasn't, but certainly Yutaro and a lot of our team are just in these circles of Degens who love to spend time on this stuff. I think just talking to maybe three to five people usually get a pretty good representative sample, just keeping in mind that everyone carries their own bias and not necessarily trying to identify Humphrey's particular needs and build exactly for those, but more to identify the larger patterns, and also to practice the general design research practices of encouraging the person to speak their mind, encouraging the person to say critical things. I do avoid, I guess, interviewing usually first-degree connections. Someone you're a little bit too good of friends with, I think can affect the results, but someone who is mostly a stranger, but is a friend of a friend is usually fine.

Jisi (09:33):

Yeah. I think that really makes sense, and especially when you're building a product that you yourself use, you can do some testing through people within the company or people that are friends of friends, and so that's pretty nice. I'm curious, how have you been able to instill this design culture within Orca as a whole?

Ori (09:55):

Yeah. I think that question is actually funny to me in a way, because I think everyone at Orca embraces the ethos of user friendliness and it's something that they're very passionate about, but actually nobody else does design at Orca. I'm the only person who does any of the design, and so in some ways, it's actually a design monoculture, I guess, because it's mostly me, which in some ways, is shocking to me that I could just design this thing and put it out there and people would love it, because it hasn't gone through endless rounds of design review, and there isn't a design boss telling me what I should tweak, which is actually very new to me.

Ori (10:34):

I guess what the culture that I've tried to instill, which I think is, you could call it a human-centered design culture as well, is just being accepting of our mistakes and this culture that everything is a trial and error, which I think I really learned actually from my engineering manager at this startup I worked at in New York called Button, where everything he'd do, he'd be like, "Okay, we're just going to try this. It's a two-way door. We'll put it out there for two weeks, and if it doesn't work out, we'll just go back the other way." I try to instill that in everything, and when I share designs with the company, I'm not like, "This is the right way," but I'm like, "This is what we're trying."

Jisi (11:11):

When you get that feedback, is it mostly qualitative or quantitative? How do you gauge where to go off from those initial designs?

Ori (11:23):

I would say it's mostly qualitative in the past. We're getting more and more quantitative in our analysis, starting to look more at things like daily active wallets, starting to look more at geographic breakdowns, et cetera. But one of the things about DeFi is that there is no KYC, so it's impossible to know in a very precise way what your users are doing. I think for a giant product like Facebook, it makes more sense to get really granular with things like AB testing. But honestly, I think what's carried Orca the farthest is leaning very heavily on intuition and that's something that I believe we'll continue to do.

Jisi (12:01):

Yeah. I would imagine it'd be hard to lean on data as well. There's a lot of spikes and changes in those patterns that it's hard to attribute to certain things. More recently, today is the number four exchange across all chains for total volume.

Ori (12:19):

It's mind blowing, right? We're-

Jisi (12:21):

Yeah.

Ori (12:21):

... what is it, 15? Not even 15 months old and kind of up there. We hit number three, I think, a day or two ago as well, and that's like flipping with Uniswap and PancakeSwap. For someone like me, I couldn't tell you what Ethereum was three years ago. It really goes to show that the space has changed, and I think it's starting to embrace a much wider audience.

Jisi (12:44):

It's so amazing, so definitely congratulations there. Also, just daily active users, like it's crossed 100,000 thresholds, so that's amazing as well.

Ori (12:54):

Thank you. Yeah. A lot of people, they're asking me, "Where is this spike coming from?" To be honest, in some ways it's like, I just tweeted about this recently, but it's UX in a different sense of developer experience, because I think we've created this SDK that developers find to be very user friendly. We've created this brand through our communications, through our community managers of having great support, and those things are also part of the user-friendly package that builds this brand that people ultimately trust and want to use, including other apps that integrate Orca under the hood.

Jisi (13:29):

Yeah, that's definitely true. I think developers are definitely a huge subset of your customer base and providing a good SDK is great design for them. It just adds the flywheel of more users finding Orca through DAP or more adapt or learning about Orca. You mentioned the brand, I think there's a lot of things that you've done really well like having the Magic Bar and making sure that you're showing that this is the best price, just reassuring users, so that's definitely building a lot of brand loyalty there.

Ori (14:01):

Thank you. Yeah. It's the most fun part of building a company for me, I think, is actually just building the product that, again, people are using every single day, regardless of whether they know that they're using it or not. Whenever I'm asked, "How did you come up with the Magic Bar? How did you come up with the Fair Price Indicator?" It really was just research either observing people using other AMMs. I remember watching people price compare when using Uniswap and saying, "Why are we making people do this? The point of technology is to automate the things that people don't want to do." So whenever we can do that, I do aim to include that in the product.

Jisi (14:42):

Yeah. That's definitely really interesting, because I think if you were to ask someone, they might not have told you that they're doing all these price comparisons, or ask for that feature, but being able to observe them directly, you were able to hone in that they were doing that comparing across prices. Branding it too is super interesting as well.

Ori (15:03):

Yeah. Branding, I would say, it's like a hobby. It's my favorite part is naming things of like, "All right, what are we going to call our liquidity mining program? Well, it's a little farming. It's all ocean themes, so let's call it Aquafarms." It's like that's my that's my hobby is coming up with names for things, but it's such a pleasure to actually do it, right?

Jisi (15:24):

Yeah. Well tell us a little bit about your new product Whirlpools and concentrated liquidity.

Ori (15:32):

I would be very happy to. It's pretty mind-boggling, to be honest, like what on earth is concentrated liquidity, right? It's just like this jargon, but I think at the end of the day, what Whirlpools is really just more advanced AMM technology that the average person doesn't really have to understand, except that it gives better prices on trades. Now, if we're going to peek under the hood and you ask, "Okay, well where do those better prices come from?" Then the answer would be, more efficient deployment of capital from liquidity providers, which is still really jargon-y. So if we peel that back a little bit more, I think what it means is that people who are providing liquidity can choose to provide their liquidity for a particular price range. So on a normal constant product, AMM, liquidity providers are forced to service trades at any price.

Ori (16:25):

Regardless of whether one SOL is 100 USDC or one SOL is 500 USDC or 3 USDC, liquidity providers for SOL USDC have to provide those trades. But instead, on a concentrated liquidity pool, like a Whirlpool, you could say, "I'm only going to provide trades when the price is between 100 and 120." Then, if you imagine on that curve before where your liquidity was spread out all across the curve, now all of that liquidity is concentrated in just that 100 to 120 buckets. These buckets in these price ranges are little mini AMMs. So each of these little mini AMMs has more liquidity than if it was across all of them, which allows this liquidity to be more impactful when it comes to providing better prices on trades. Hopefully, that makes sense. I know it's a little abstract.

Jisi (17:17):

Yeah, definitely. Seems like you get a much better yield and you have a little bit more leverage with the amount of capital that you can provide by choosing that range. How does Orca help users determine what's the best range to choose?

Ori (17:32):

Minimally, I would say. I think this is one thing where we've taken a deliberately moderate path. We don't mandate the price range. We do provide some presets that people can select, because one thing I observed from users on Uniswap v3, which is, of course, the OG concentrated liquidity AMM is that many people, even very sophisticated folks would just get to the point of choosing a price range and then be like, "I don't freaking know," and then just give up. So I think even just giving someone like a reasonable default, it might not be the most optimal, but a reasonable default is pretty good. So we give defaults based on things like volatility over the last 30 days, or over the last X days, which are generally a reasonable place to start. Then we give a very lightweight suggestion of, "With this kind of price range, you'll want to rebalance your position more often, or less often." But ultimately, the folks who are very competitive about it are almost always going to go custom.

Jisi (18:30):

Yeah, definitely the presets and having a recommended thing from Orca is definitely really useful. Obviously, concentrated liquidity is something really complex, but Orca is making it more intuitive to understand. How do you find the sweet spot of that kind of abstraction when you're catering to your users that may be more crypto native and some of them may be new to crypto, how do you find the right point to extract complexity and simplify?

Ori (19:00):

This is where we come back to that discussion of, "Who's our target user right now?" Our target user for concentrated liquidity is not my dad, very sweet, still doesn't really understand what Orca is; knows it's a coin, not much more. We're not trying to convert people like him to become liquidity providers right now. So when we are thinking about something like Whirlpools, it's folks who maybe have a bit of experience with liquidity provision already on a constant product AMM, all the way to people who have already been very sophisticated market makers on order books, or on concentrated liquidity AMM elsewhere. It's from there that you can imagine, "Okay, what vocabulary does this person who already does constant product AMM provision have? What do we need to teach them in order to get them onto concentrated liquidity?" That's where we have a progressive reveal approach in our UI. Not everybody loves tool tips. I fully admit that they're not elegant or beautiful, but they do serve a purpose of allowing someone to gain extra information when they want it without cluttering the UI.

Jisi (20:08):

Yeah, definitely. I think that reassurance and that extra education is pretty important, but it's nice to keep things simple as well. What are some of the biggest challenges you see in crypto in terms of user experience?

Ori (20:21):

The biggest challenges I see, well, I think there's two. Actually, one that I'd love to highlight just because we talked about branding already is just the branding of crypto. It's terrible. It's so off-putting, I think, to the average person, like very sci-fi just very complex jargon and terms and mining and proof of what have you. I think all of these very complicated terms and all of these very technology-oriented terms can be very scary for people, and so just overcoming that. Even the name crypto alone, think of the term like digital money versus cryptocurrency.

Ori (21:01):

Cryptocurrency is so much scarier, so I think first just getting people over the hurdle into thinking that they're smart enough to use crypto is one thing. Then second, I'd say DeFi has its own set of challenges. Compared to something centralized where there's KYC, there's usually more of an insurance fund for users. With crypto, it's like you're just thrown out into the water and you better swim. If you lose your private key, it sucks to suck. So I think creating a more gradual bridge for people is a huge challenge.

Jisi (21:34):

Yeah. That makes sense. It seems like crypto needs to hire its own copywriter or someone in branding, and then also just more education, more, more on ramps to becoming full Degen instead of learning by yourself and making those huge mistakes, but ones that you do remember.

Ori (21:54):

Yeah. Even the term Degen, I think, is very telling. It requires such an investment of effort to learn to use this stuff.

Jisi (22:03):

Definitely. Yeah. I have a question, there's a lot of designers that are in Web 2.0 and they want to crack into Web 3.0. That was actually one of your points in your op-ed in, in that there are a lot of design patterns or design decisions that were pushed forward by Web 2.0 that could be brought into Web 3.0. I'm curious, what kind of advice would you give to a designer that's looking to crack into Web 3.0 That may potentially have some Web 2.0 experience building?

Ori (22:31):

First is to, of course, learn about crypto, Defi, Web 3.0. People will find it a lot easier to onboard you into their Web 3.0 company if you don't have to ask all the questions about the basics, and there's tons of free resources about this nowadays. I'm part of a couple of communities online that really target towards beginners with their educational resources. Your experience as a Web 2.0 designer is incredibly valuable, and I think anyone in the crypto space knows that crypto companies desperately want to hire designers nowadays, but just can't necessarily find them. So I think if you come with a clear interest in breaking into Web 3.0 and also with just a little bit of due diligence, knowing the basic concepts, you'll be pretty much welcome with open arms.

Jisi (23:18):

Nice. I'm curious, looking ahead, what are you most excited to build at Orca?

Ori (23:24):

Good question. Yeah. I've spent the past couple of months, obviously, building Whirlpools, which is this super technical product. Looking forward, I'd love to spend some time building a much less technical product, or at least a product that looks a lot less technical on the surface; something that could potentially black box some of the complexity and open back up some of these opportunities to people who are maybe just starting out, or maybe just a little bit nervous about dipping their toes into crypto. I don't have very specific plans to share yet, but that's definitely where we're thinking nowadays and I feel like we're in a good position to potentially build out something very different than what you might normally see in Defi.

Jisi (24:13):

Nice. Well, that sounds super aligned to the Orca brand and seems like a really good place for orca. Also, dipping your toes is probably a good tagline with the whole ocean theme.

Ori (24:26):

Yeah. One day I'm going to run out of ocean puns, but I'm not done yet. 

Jisi (24:31):

Speaking of ocean theme, could you talk a little bit about some of the climate impact work that you're doing with Orca?

Ori (24:38):

I would love to. The Impact Fund, it's like my child, I suppose. I just really love to see where it's grown into, but it really was the result of just a conversation that I had with Yutaro one day, because I've spent a lot of my career in the past really focused on just impact work; whether that's ed tech or trying to work in FinTech for developing countries, that type of thing. One thing I worried about with crypto is, "Is my time really being spent to make the world a better place?" I've become convinced that building out essential infrastructure for crypto is, in its own way, a good thing. Having influence in this space to try to push it in a positive direction is a good thing, but I wanted something a bit more concrete.

Ori (25:24):

When I floated the idea of just directly donating a portion of all of our training fees to charity, I was actually really pleasantly surprised to find that everybody was very much in support of it, like Yutaro, our advisors, even our legal council, they were like, "Yeah, great. I love it," and so we do do that. We donate a portion of every single trade that goes through Orca. A little bit of those fees go to charity and it's over a million dollars accumulated to date, and so we'll be donating those to charities that are focused on climate change and sustainability, which is both what I think is the most pressing problem facing the world today, and something that happens to be very aligned with Orca's marine brand. So it's something I'm very, very proud that we have the opportunity to do.

Jisi (26:09):

Yeah, that's amazing. It's definitely super aligned to the brand and definitely a great thing that Orca is pushing as a company for our closing question, who is a builder that you admire in the Solana ecosystem?

Ori (26:23):

This is totally a cop out and a shill, but the first thing that comes to my mind really is my co-founder, Yutaro. I think often I get a lot of the spotlight just because, well, I'm much more talkative than he is, and people, I think easy to talk about human-centered design and all that we're at. But he's really the brains behind the operation. And also the one who's invited us all in into crypto. When people ask, "How did you start learning about crypto?" I think back, and I'm like, "I was so lucky that whenever I had any question, I would just ask Yutaro.

Ori (26:57):

He would very patiently and excitedly just tell me all about it. He has these deep roots working for the Ethereum Foundation, and so I think those are something that was very respected. In Solana, it's not about this tribal, "It must be Ethereum." "It must be Bitcoin." "It must be Solana," but really more of this holistic look that he has into the ecosystem and where it's going to go. I think the fact that he saw Solana back in September 2020 and said, "This is the place to build," and was absolutely correct just says it all.

Jisi (27:28):

Yeah, definitely. We would love to have him on the podcast as well one day. That would be amazing.

Ori (27:35):

Yeah, you should totally bring him on. He was also voted in the share house that we live together, best voice mail by our co-living mates. He has a great podcast voice, so I'm sure you'll have a great time talking to him.

Jisi (27:49):

I like hearing him in the Orca demo videos.

Ori (27:54):

Oh, I'll tell him that. I think he'll be like, "Oh, that's interesting," but actually very happy under the hood.

Jisi (28:01):

For sure. All right. Well, thanks so much for coming to the podcast, Ori. It was great to meet you. If you like to check out Orca, you can access the Dex at orca.so and at Twitter @Orca_so. You can find Ori on Twitter @Oritheorca.

Ori (28:19):

Thanks so much. This was a really fun conversation.

Jisi (28:22):

Yeah, for sure. Thanks so much. Have a great day.

Ori (28:24):

You too.

Jisi (28:25):

Bye-bye.