Ep.40 [EN]: Cindy & David: the Driftooors building the future of finance on Solana
46:05 Share

You are now listening to Hotlong. I'm the host, Mabel Zhang. This show attempts to shed light on the broader Web3 community. The guests will include Web3 operators, investors, or ecosystem participants of other forms. The show will also have a special focus on the Asia-based Web3 world. Connect East and West. Enjoy the warmest Web3 conversation.

Hello everyone, welcome back to the latest episode of Hot Long. This is Mei Bojang. Today we are super excited to welcome the co-founders of Drift. So what is Drift? I guess we can also call it the future of finance, but technically it's on-chain perps.

I'll let Cindy and David introduce themselves first, and then we can talk about their three years of building. Happy to kick it off quickly. I'm David, one of the co-founders. Cindy and I have been working together for quite some time. We actually started our first journey prior to Drift.

building a hedge fund, trading per petules and options and other derivatives. And that motivated us to think about building an untamed decentralized version of it, which was essentially far more trusted and transparent at the end of the day. I think the ethos of crypto for us is very much open source. And it was always very strange to us why centralized exchanges and trading took place off chain. So we came in with that mission and a goal to solve that.

I'm Cindy. I'm one of the co-founders of Drift. My role is mostly in strategy, products, and BD. And we've been building out Drift since 2021 and really excited to have gotten to such a long and meaningful journey to TGE this week. So super excited to be talking to you today, Mabel. We've known each other for years as well, so I think this will be a really fun conversation.

Cool. Yeah, I was gonna dive into the conversation by talking about the three years of building. How does it feel having been through the whole bull and bear and bull again?

I think it's pretty humbling to round trip it, going from bull to bear and then back. I think there's obviously things that we can control for, but also things we can't control for. I mean, things we can control for is building the team, staying strong in difficult times. Obviously, for where we're built, which is on Solana, a lot of it has been at the mercy of the market and factors that were wildly out of our control. So I think for first-time protocol founders,

We were obviously founders in our past lives running a hedge fund, but as first-time protocol founders, I think it was very humbling to go through that first experience. I think...

When Solana obviously had a meteoric run, we naturally got a lot of attention, a lot of volume, and just a lot of traction in a very short period of time with a team of five people. And I think initially, we definitely conflated that initial market run with our ability to execute, which was completely untrue. I think that definitely got to our heads when it just ran up. And I think realistically, price went up, our volumes and performance went up.

And we, I think at the time, were almost a little too naive to figure out what noise was. And I think that got to us. And I think it was a very humbling experience to go from 100 to 0 when we did go through some challenges for V1. I think that took place with Luna.

you know, the Loon incident took place, we found ourselves at the mercy of an accounting bug. And I think that really set, I think what is now Drift's DNA, so we sort of refer to it as like V1 and V2. I think everyone that's since joined on, and this is something we're very proud of, is not a single engineer has left our team since that V1 incident. I think everyone stood strong. I really dug deep in a really difficult time. Obviously we had our own internal challenges, but then we also had challenges with FTX,

There were a lot of doubts, frankly, around Solana's future and what that might look like. So I think to really push back against those odds was really...

I think warming for us as founders to have team members that still believed in the mission. I think it's very easy to look at alternatives when things are tough. So I think that mission has kept everyone really much glued together. I think obviously that bet paid off a year and a half later where we started to see Solana really come back. Drift came back. I think it's one of those things where 90% of it takes place in

without any recognition of credit, of success, and it all happens at once. I think what people see today is actually a hallmark of three years of work, not just the past cycle. So I think that's definitely been a big lesson for us. I guess I was curious to maybe for you to share a little bit more about the design for V1. Now it's like really long time later, but it will be interesting to hear from you how exactly the incident happened.

Yeah, with V1, our matching algorithm was a pure virtual AMM. What this means is that all the positions were PVP. So traders would take positions against each other and there was no liquidity backing the VMM. And around this time, it was a very popular model for bootstrapping perpetual markets back in 2021 because it meant that you could spin up

You could spin up perpetual markets with 50 to 100x leverage with no collateral whatsoever, except for the collateral of the trader before you. So it relies heavily on arbitrage mechanisms to make sure that price stays in line with an external source. The trouble with this mechanism happens when a risk layer is not introduced between the virtual AMM and the settlement of positions.

So what happened was that there were too many on the day that Luna went from $210 to $0. There were a lot of traders arbitraging the short leg on Drift while getting long on other exchanges. So what happened here was that there was too much on the short leg that people who went long were not liquidated fast enough.

There were simply not enough people who went long, too much imbalance in the market. So this group of traders made too much profit and couldn't cash out. And the way that V2 works now draws upon years of experience of risk management in the industry, as well as learning deeply from this incident, where we now have a P&L settlement layer that's built between the exchange and the

and the matching algorithm where you now need to make sure that each loss in the protocol matches with a win on the protocol before they can be withdrawn. So this would avoid the mistake where if there are too many shorts on the exchange, you kind of are not able to have enough profit in the platform. So this protects the exchange and the insurance fund from being able to be drained essentially.

And there are a lot of other things in the matching algorithm outside of pure PNL settlement that makes V2 way more robust than V1.

So V1 was pure virtual AMM, but we recognized very early that pro traders need an order book in order to get the fairest pricing and the best pricing. So in V2, we built a order book plus AMM plus an RFQ matching algorithm to enable external market makers to come in, provide collateral and liquidity for traders, and this means that

each position is fully collateralized by liquidity in the platform versus with the virtual AMM that sort of acts like a provider of last resort without the necessary collateral backing it. So now most positions on Drift are actually over collateralized because there's a lot of liquidity providers on the other side. This means that even though you're taking 10, 20x leverage, you're still supported by a lot of collateral on the platform.

So how did you actually power through the incident? Because it sounds like you had a really hard time. I think we caught up right after the incident, but David didn't exactly tell me what you guys did. So I was curious. I think we...

we obviously, I think as a team got together and made some pretty difficult decisions. I think very fortunately for us, this came right around the time of our series. I think I mentioned that we had a lot of great success, a lot of great traction initially. So we did get a lot of support from investors and,

I think naturally for investors, the fact that $15.5 million was the whole and something that we took directly from the series and had to plug that was a big decision. I think it was very easy for team members to walk away from that. But we knew that in order for a successful V2 to take place, we couldn't operate with that kind of weight for the team. And I think that also wasn't in line with the values.

And I think that's really what glued the team together. I think a lot of the early members collectively came to that decision that in order to do our best work for V2, we had to move forward without any types of constraints, without any kind of debt, both in the physical sense, i.e. we owed users money, but also just like the mental debt of having to assign, you know, wake up every day,

with the potential anxiety of people in Discord being upset with you. And it was, you know, a big distinction at the time. We essentially were on, we essentially underwrote ourselves and underwrote our own risk.

gave ourselves a fairly fixed window to execute on what was a very ambitious vision. I think that collectively enabled us to stay nimble, I think, and just stay efficient as an organisation. We never hired more than who we needed to. I think, and it really forced us to think about the first principles of what are we doing here?

and what is the most effective way to get from A to B. And I think that's what you saw with Navitu. There was a lot of building, there's a lot of obviously testing and solving of immediate issues. But I think building a platform that was innately just more scalable than the initial product was a big part of it. So I think that

ultimately united the team. It's obviously very easy to say in retrospect, but at the time, it wasn't so easy of a decision. I think there were a lot of different ideas being floated, but we knew that...

I think that was a turning point for how we defined ourselves as a company and unfortunately that's what's become the Drift DNA to date. Very impressive. Did you share this story at all in the past? I think this is the first official, I guess, voice version so I think that's hot at the press for you but

But it's one of those things where we wanted to let the results do the talking. I think it's obviously easy to make these kind of claims, but if we made these kind of returns and we weren't able to back it up, I think that's not in the nature of how we do things here at Drift. We want us to show a great product.

you know, have all these features that other, you know, other competitors may not have considered and really let that show how we've gone and gone past what we started. So I think a lot of why the story had a bit of a time delay was, uh, I think we, we, we wanted to just, uh, we wanted to stay low, make sure that,

things got done versus going back and forth and creating more of a target, which again leads to more noise. I think I was talking about streamlining and efficiency. All these vectors present additional noise. So to the extent that we minimise that and we focus on building and getting from A to B, that's the decision that we chose. So I think we essentially cut off contact with

a lot of external media stories for a long time just to make sure that every possible thing was done to get V2 over the line in as short a time as possible. You know, the biggest lesson for the past three years for me was that you just never get off the table and usually it will result. All right. So Cindy talked about V2 briefly, how it compares to V1. My first question about V2 is that can you define a participant of the ecosystem in a few categories?

Yeah, for sure. We largely break people down into four categories. We have traders, we have liquidity providers that provide liquidity on the EMM. We have market makers who provide active liquidity on the order book, as well as the just-in-time auction. So every time a market order comes in, market makers can choose to provide RFQ liquidity, and that's all active.

And then we have order book keepers. So our order book is unique from other on-chain order books because the entire mechanism is held on-chain. You have a network of keepers where each keeper functions like a solver. They solve and match orders together when their bits and odds cross and place it back on-chain. So this helps to maintain efficiency while keeping things transparent and fair on-chain.

And so, yeah, this is the four large group within traders. You have all sorts of traders on the platform. You have people trading small size to huge whales depositing capital on the platform. And on the liquidity provider side, we would count depositors in that group as well. So anyone who's staking insurance in the insurance fund or stock.

depositing capital to run basis strategies, etc. So yeah, large, most of our users fall in the first category, which is active traders. I mean, you only started with VAM, but then later on, you decided to offer spot trading and swap and exchange with leverage. And you also introduced a money market protocol that allows and then allow people to stake in the insurance fund. What was the thinking process behind it?

Yeah, for sure. I think a big part of building Drift to be a fully decentralized platform is for elements of the protocol that I think are quite keystone to it to be eventually held by the community. So I think this was a big mission where we had seen examples of big insurance funds like BitMEX.

obviously be a backstop to any insolvencies or bankruptcies that took place. But also as another revenue stream, I think that insurance fund is still today one of the largest. And it's essentially made up from just a portion of the fees being sent there. So we thought if there was initially it was a way to bootstrap.

alongside plans to decentralize this, you know, could there be a way for users to get a part of that, uh, part of that yield by backstopping the protocol? So as part of, uh, starting off, we were thinking a lot of users would definitely take the position of, I guess, the exchange. If you think of the exchange as, um, uh, like a house essentially, uh,

People are coming in, trading in and out. So it was a good way for traders to essentially represent some support and underwriting the risk that protocol would otherwise underwrite itself. So I think it was really crowdsourcing and democratizing what that looked like.

And I think the insurance on product was the first of many products that then came around. I think we can touch on DLP, which initially started off as Drift Liquidity Program, now renamed to BAL to be more reflective of the type of product it is. But that is also a product that we came up with in a way to decentralize market making evolves as well.

taking inspiration from the GMX, GLP predecessors as well. So I think ultimately it comes down to can we get more and more participants from our ecosystem to take on key functions of the protocol so that

at some point in the future, I guess the core labs team can step back and the protocol has enough participants in each of these respective areas to essentially carry itself forward. So I think ultimately a way for the protocol to decentralize far more effectively than if we as a centralized team were to control each individual component of it. As Cindy mentioned earlier, initially,

The challenge with the VMM model as a standalone was essentially an imbalance in the system. So what we thought of was if there was a way to plug that and have some type of representation on the other side, in the form of some type of LP staking, that would be the way to do it. So if you were to go long and then in the VMM model the protocol goes short, one could represent in a way that

users would be essentially taking that short position or essentially the converse position to what the users were taking. So in particularly choppy markets, we tend to notice that users typically are on the losing side. They typically get chopped up. So products like these do very well in crab markets.

And when we launched this product, Sol and a lot of assets were essentially ranging quite strongly. I think that represented a good way for users to essentially express a certain position in that market. Obviously, as markets became more directional, this type of product can get quite risky because essentially,

It's a single direction, so that's something to be mindful of. So think of DLP or BAL as almost like a base layer for users to participate in. And then on top of those, what we observed was interesting was other teams come to Drift to build out more sophisticated vaults, whether it be hedged DLP or BAL vaults, where you would essentially take profit on certain legs. So I think the development of those, I guess...

call them like gen 2 primitives was really exciting to us where people thought okay as a stand learn there was enough opportunity to get involved in providing liquidity to the system and then people came to us to build additional products on top of it so i think when we saw the emergence of those teams we got really excited because again it was no longer just the team pushing this out but rather uh the community taking it forward i think as it stands it's

Cindy correct me if I'm wrong, I think it's close to 80-90% that is essentially community funded. Very little actually comes from the team. So in terms of the proportion, it's very minimal input from the team and large from our community. So how's the market making process look like on Drift? Can a retail participate in that process?

Market making on chain is a pretty interesting topic to talk about because it's quite different from centralized exchanges where centralized exchanges give you an API to integrate with

on-chain exchanges and the order book, there's a lot of difficulties and the intricacies that come with your on-chain market making. So instead of going through an API, you have to go through sources like, you know, going through a Rust SDK, a Python SDK. And what we initially thought was, okay, it would be quite difficult for MMs to integrate

But what was really bullish about our launch and our community is that we saw countless of community bots being spun up. We saw people launching vaults that allowed you to market make on behalf of other users who are not native market makers. And then we saw, especially during the big Solana run up last year, market makers who are more on the institutional side start to integrate

more closely with Drift. And so initially what we consider to be a difficulty barrier became pretty clear that because the platform was fully on chain, there was clear utility for using it and clear taker demand as well that pushed makers to get over the initial bump to integrate. And now it's way easier to do it because we've also had community members

help with building SDKs and building ways to make it easier for MMS to integrate. It reminds me of Solana back in 2021, where there were not many docs on building on Solana, only maybe five devs knew how to deploy on the chain. But so many of the community came together to form documentation guides and bringing more people in that we're seeing a similar energy now.

And with community, we've really tried to emphasize having people be able to deposit and not having the technical skills needed to run an MN bot. So we launched a couple of community market making vaults that allow anyone to deposit capital passively into the protocol. But that capital is being used to provide liquidity.

And we found huge success in this. Some of the walls, I think, in aggregate have about 100 million in TBL.

and are responsible for 30-40% of the liquidity on the exchange. So yeah, this is a really good win for the community. I noticed that on your dashboard you have protocol-owned holdings. How do we understand that? Protocol-owned holdings are one way to think about fees on the protocol. The reason we call it holdings and not fees directly is because protocol holdings are made out of a couple of fee sources.

and that includes trading fees. So Drift takes a trading fee from all the Perp trades on the platform and we've actually been running a couple of promotions in the last month to make us the cheapest place to trade Perps on-chain just in anticipation of TGE which has seen a really really good and solid reception from the community.

And it also includes, because we have an AMM owned by the protocol, liquidity fees and funding rates earned by the AMM. So it includes not just trading fees, but activity done by liquidity providers.

And thirdly, probably one of the more the larger revenue sources for us is liquidation fees. So with liquidations that happen on the platform, the protocol as well as liquidators share liquidation fee from

from users getting liquidated. And that's part of why I guess to preempt something in the future that we can talk about is that for the airdrop, we prioritized users who got liquidated through what we call a rec drop because that's probably the best anti-civil measure that a user can have.

And earlier you mentioned a situation where there's more unsettled P&L than what's available in the pool. What exactly happened when that situation is there? The main thing is that unsettled P&L has an asset weight of zero. So this helps to avoid incidents with what happened with Mango back in 2021 or is it 2022?

where a single trade on an illiquid market was able to push up the price of mango, I think 10x within the span of hours. And this was done with manipulating the price oracle of an illiquid asset.

So what happened there was that the unrealized P&L of that position was used to collateralize other positions, which enabled the attacker to take up money from the protocol. In our case, the unsettled P&L layer prevents this from happening by making sure that that unrealized P&L, even if a market is being manipulated, doesn't allow your funds to be withdrawn from the exchange because it has an asset weight of zero.

So you need to settle before being able to take your funds off. So the only thing that happens when there's not enough P&L in the market for a settlement to happen is that you need to wait for the P&L pools to be replenished. Because I guess on a PERP exchange, you typically have more losses than winners. It doesn't happen very often. Our P&L pool is usually at a very healthy level compared to OI.

And so we haven't had an incident where this is the case yet. People learn a lot from that previous incident. So I think this is probably the safest for now. On your perp market, you have quite a few assets. How do you pick which assets to start a market? I think this depends on, I guess, the stage of our journey. So I think to step back a little bit of time, I think there are some core assets, which I think every perp deck should have. And I think that was

very much a default and necessary type of listing. So we think major caps, major coins that are traded, top 10, top 20. I think that is, I think parity is sort of the first thing that we strove for. But beyond that initial phase, a lot of it is thinking similarly to any other exchanges, how can we be the fastest to list certain assets that are taking off? All you notice is that in the first 48, 72 hours of trading,

of a very interesting asset that the community rallies behind. Those are the coins that we want to start moving on ASA as soon as possible. And I think it really is a speed game. I think in a game of attention, really. So if you are the first to list any asset, I think it's an identity that we've tried to craft for ourselves is the first place for certain perps, particularly Solana assets. We've done that with things like Bonk, first Bonk perp, first for WIFI.

And a bunch of others as well that I think have, well, we've always tried to be the first on Salon to make that happen and give traders exposure to these kind of assets. So I think all we need to do to list a market is really an oracle. So it's actually designed in a way where the listing itself is fairly easy to do. Now, I guess the curation of what that looks like has a lot more intricacies behind it.

I think a lot of it, we try and get that feedback from the community. I think narratives are a constant thing in crypto. I mean, recently AI coins have been big. So I think we've correspondingly listed certain coins to reflect that as well. I believe Tau and a few others are on Drift, or I think not on Drift rather, but they were essentially shortlisted by the community. And then from a shortlist, we end up listing, I believe, Tau as part of that. And we've had previous...

discussions around meme coins, for instance, we also have things like Pepe with things that might not be on Solana take place on Drift, which is obviously on Solana. So it's really giving people the opportunity to jump on narratives,

in a way that they can't find elsewhere or natively. So I think that's what goes into that. Now, then you might also think about risk as well with some of these meme coins that have low liquidity. Obviously, to what Cindy was speaking to earlier, there are essentially risk vectors and you don't want

a single meme coin market to blow up everything else around it so we have essentially what we call isolated pools where it essentially has like a limited insurance fund so only the fee pool is used to protect against levered losses and we made that very clear so it's essentially striking a balance between risk management but also embracing some of the more creative degenerate type of coins that users like to see and get excited about so we want to make sure that we're not

taking any, we're not making any decisions to harm the longevity of the protocol, but also doing it in a way that gets people excited and keeps us relevant. So I guess we take listings quite seriously because whilst listings can be easily done, we don't want them to be obviously abused. And a lot of it now at least comes from the demand of what the community looks for. So as part of our next stage and post TGE,

essentially letting token holders vote on what they would like to see listed. And I think there are quite a few models that have been already discussed, whether it's one market a week or people can essentially push proposals through to list really exciting coins that they might be noticing on other chains, but still don't have a way to get access to on Solana, for instance. So

permissions listings to get quite exciting is if people in the community take these exciting listing ideas to us sometimes coins that we might not be aware of so long as we have an infrastructure i.e the oracle to make that happen it's a great segue for us to push that because you're essentially crowdsourcing uh reach in ways that you couldn't before so we don't want to be the centralized authority pushing listings we want that to be very much democratized

Do you also follow any like Twitter metrics analytics or I guess like any sort of metrics analytics for yourself or is it just mostly relying on community curation?

I think to an extent, and Cindy, I'll probably let you jump into, we keep a close ear to the ground on things outside of Solana. But obviously with just a small team, there's only so much you cover. I think we do run some somewhat rudimentary Twitter-based just algorithms, bots, just to see what's been mentioned, what's been referenced. But obviously those are very much...

skewed to the type of material that you follow as well. So I think obviously there are biases and indexes to the type of coins. I mean, I think both Cindy and I are likely fairly active in the degen circles in Sydney probably more so than I am. So I think that influences what we see, but I think we're trying to get more clever around how we think about what to bring into

to Drift that is exciting and relevant. But I'll pause here Cindy, feel free to jump in. Yeah, we definitely try to follow what people on Twitter, crypto Twitter especially, are trading. So yeah, looking at big mentions of the week, narratives as they form, because I think one of the big parts about being a PerpDex is making sure that we always have the asset that people want to trade.

So, yeah, being the first to list a pre-market, which we launched earlier this year, that enables people to trade markets that are not live yet, even across chains. So allowing Solana users to trade assets that are outside of Solana has been a pretty powerful outcome of the product that we've built. So, yeah, I think all those things in aggregate inform us on those things.

I was curious if you know your demographic look like? From how we think about it, it's how what we see, we obviously don't track things like IP, but based on how they trade, we've almost said that 98, 99% are retail, just based on how markouts look like. So when a retail trades, typically the type of behavior is market orders. And they're almost, I wouldn't say completely price insensitive, but they're not,

essentially submitting the orders in a clever way. So I think this is where market makers come in for. So I think based on that activity, you can essentially tell what type of trader they are. I think for us, particularly in the last six, seven months, we've noticed that 98% of these guys are retail. So it's really, again, power laws around. There's a small bunch of users that obviously do a lot, and those I think we could categorize as institutional. But what gets us excited is so much of it is just very natural take-a-flow.

And that's probably something we're most excited about. Separate to the exchange metrics, we're getting a lot of community groups come from the likes of Turkey, Korea, and even India as well, trying to form their own communities and get excited on how to best interact. I think there are so many, there are, I think, recently how-to guides from these different regions.

that we just didn't see before. So I think part of a post-TG plan, thinking of diving deeper into these region specifics, and it even might even be something like

you know, having our website render in Korean language and UI to represent that accordingly, such as, you know, red is up, blue is down type of thing. These are localization strategies that we haven't thought about, but I think once again, very critical so that any local demographic can adjust for what their needs are. Because right now it's sort of a one size fits all. So I think very surely we're going to move away from that and evolve.

How did Solana congestion impact your performance and what did you do to try to optimize experience for the users? Solana congestion was particularly bad I think about three or four weeks ago now. I think that was one time where a lot of the market maker transactions were failing.

Interestingly, because Drift has optimized a lot for minimizing compute units for each transaction, we were less affected by this than other DEXs. We were still getting pretty high rates of transaction landing. But as a part of that, I remember we had a big one-week sprint where everything was about maximizing the chances of landing on the chain. So...

we experimented with a lot of ways. One was with dynamic priority fees. So we launched a product called Trader Profiles that allows anyone to customize their priority fees that they pay on chain. And if they want to increase their priority fees, they can do that. One, I guess, common criticism with priority fees is that it doesn't tell the whole picture. Priority fees may

only increase the chances of you getting filled. But because of how the scheduler works right now, it doesn't guarantee inclusion. So you end up paying a lot of fees while not getting filled. One way to get around this is using JITO bundles to submit bulk transactions at once directly to the validators. So we started experimenting with allowing JITO bundles for our MMS and for our transactions that went through.

And that improved build percentage by quite a lot for us and users were very happy with the outcome. The other thing that was pretty critical for us is that we had multiple RPCs during that time. So users could pick between Triton, Helios, and a couple of other RPCs that are the best in Solana.

And what this means is even though one RPC is congested, you could go and swap out for other RPCs. So there were a lot of customizations on the UI front that enable a better experience for users so that they're not just stuck, you know, not being able to land a transaction. They can actually either pay more priority fees to choose how they want the transaction to land.

Ultimately, I think Solana's congestion issues are sort of fixing themselves with the latest updates and the new scheduler that's coming up. We're really hoping to see this improve over time and as more scaling solutions hit Solana too. I didn't realize that you guys actually did some optimization and that was fast, like within a week. Yeah.

Yeah, it was like nothing else mattered that week besides getting people. Very impressive. Now, can you talk about the drift token distribution for the users? I think people are curious about this. And actually, I had a community collected question for this. People were keen to understand what's the thinking behind not considering trading volume and points accumulated in distribution terms. So the overall thinking for the airdrop was that

We wanted to, obviously the protocol has been live for over three years, so the points program itself was only three months. So we had to be very holistic with how the airdrop was designed because they had to reward everyone who participated in the protocol from day one.

And so a big portion of the airdrop went to what we consider OG users. So everyone who came in and used the protocol prior to the points program. Points program was counted separately and was designed to incentivize a couple of things, not just trading volume, but amounts deposited in the platform and ensure that the retails who weren't able to push a lot of trading volumes still got a minimum amount of drift.

for governance and decentralized purposes. We wanted to make sure that everyone had a good amount of tokens. It took a lot of careful thinking around distributing between so many user groups. I think in aggregate, we had over 200,000 users. And so thinking about where things should go was mainly made for the future of the protocol. It's like, who are the real users of the platform?

who are going to stay. So we looked at things like unique days interacted with the protocol. So consistency was a major factor in what we had considered. And the other factor was, did people stop using the protocol when the points came out?

Or when the points ended, or were they really using it because they liked using the platform? So there were a couple of tells in the data that helped filter out real users versus civils. And that resulted in what we consider a very fair airdrop allocation compared to pure linear approaches.

I saw people complimenting the design though. What was particularly liked by people on Twitter? I think we can break down to a few things. I think one is that we had a maximum period of three months that we set. So we weren't going to go beyond that.

I think there was a lot of airdrop fatigue where people were getting farmed for season on season, season one, two, three, four, and the list goes on. I think at some point people were just fed up with that. So I think every subsequent points program was very much diminishing returns for users. So we from the gate said that we were only going to do three months max. We could shorten it.

but we weren't going to go longer than that because the history of protocol is longer than three months, it's three years. So I think immediately that timescale was very important to us. So we said that very clearly and we stuck to that. I think the second was thinking about

uh how to distribute the points uh we didn't want a situation where a ton of people could get a lot of points in last week by virtue of just trading large size or deposits in large amounts so to circumvent that we allocated a fixed amount per week uh to people so that you know there wasn't it was less likely to be game i think fell more in line with consistency

I think consistency obviously underpinned the groups that we were trying to target, people that were trying to use Drift and not just come in and then just withdraw. Obviously, we didn't penalise withdrawals. I mean, we understand that people need funds for different things, but those are some guidelines for how we're thinking about indicators. And then finally, I think the whole OG points, a lot of people had given criticism for just Solana Project in general for...

almost over waiting this points window and I think for us we were very deliberate to keep OG points hidden for the entire period so people didn't know exactly where they stood with OG points but if you were just using the protocol on a consistent basis you were handsomely rewarded so I think by not

putting that number out there uh it did two things obviously there was like some element of fun speculation people were like how much should we get um but i think more importantly we wanted to make sure that the people who mattered the most to us were getting the largest i guess pool so i think it ultimately ended up being like 64 went to ogs and then a smaller amount went to the points program and then i think other parties like keepers were also rewarded too

So I think when it came down to the overall breakdown, we felt that the balance was very fair. That's, I guess, just on the point side. And I think specific to the airdrops, we had this clever mechanism that our engineering team came up with, which was to essentially have a six-hour delay claim period, where if you claimed six hours later, you'd get a 2x on your media allocation. So that 2x would be a bonus.

I think, again, it's trying to identify and essentially get rid of the Sybilers, people who would essentially claim and dump immediately. I think the botting software has gotten so easily accessible by a lot of people where you could see just red candles on the first day. And I think these are the users we want to flush out. And I think...

Naturally, if you're running bots, you can essentially claim it and sell it. But again, you wouldn't know that you only got half the amount. So again, in the spirit of rewarding the true users who, again, six hours isn't a long time if you've just been consistently using it. I think it was a very easy litmus test for us to figure out who was going to keep using the protocol, using the platform versus who was very eager to just claim and flip. So yeah.

I think the community appreciated that because it was probably, it was the first time that a team had thought that level deeper on how to make sure that their rail users were protected and looked after. The claim window, I guess, on the final point, the initial checker was a good temperature check for how we went. I think, you know, we obviously didn't

The feedback was largely positive, but we took that as another indication of what we could do better. In the week between when we first dropped the checker till today, we identified more simple addresses and as a result, sought to do a redistribution accordingly. Again,

trying to strive towards excellence and it's never going to be perfect. People are always going to be upset and there are always, I guess, false positives. We might penalize a person who was doing legitimate activity. But again, having the right channels so that these people are still covered is important to us. Right. So the long term users and the loyal users, basically.

We come to the last question of the pod today. It's a very generic one. So what's on the roadmap for the next six months? And specifically, have you thought about a mobile app? Mobile app has been one of the more highly requested apps

from our community and also our team. People are very excited about being able to bring their trading with them. And not just mobile apps, but I think different ways of interacting with the protocol. So Telegram bots have been super popular this year. We are, I mean, no one has launched a perpetual based Telegram bot yet. So I think this is actually one of the extensions of mobile where you have a wallet, you know, custody within Telegram

or even self custody, like what Dialect is doing. You can trade anywhere you go whenever you see a Malpha SS here, you know, out and about. And I think this is a pretty powerful tool that we're hoping to explore more into in the coming quarter.

Other than that, one of the big themes for Drift this year is focusing on asset issuance and working with institutions on bringing more TVL on-chain, especially through real-world assets. So one of the cool things about Drift, and the only decks that you can do this on, is cross-collateral assets.

And cross margin has been a hugely, hugely favorite feature from traders where you can essentially use any asset as collateral on your perps, not just USCC. So in a bull market, you get to hold on to your sole position or GITO sole NELST and be able to leverage using this as collateral.

So imagine taking this concept and extending it towards building perps that are collateralized based on any RWA like BlackRock's Biddle token.

that maybe sits in a permission pool or FTX estates, LockSol or even going further than that commodities and other types of assets that you don't normally see on chain. What we're really interested in is building a harbor on chain where you can deposit these assets and use them to build structured products around it. Think of it as more of a generalized asset management platform.

a place that you can use to find yield in different ways rather than a pure perps trading platform, which is why we've been kind of expanding the boundaries between a simple perpdex. Because I think as that space gets more competitive, you'll need more reasons to attract users and depositors in the platform. And I think yield ultimately is still the best way to do this. If you think about innovations in the perpdex space, funding rate as a instrument is also important.

It's pretty interesting where you can tokenize, if you can tokenize the funding rate, similar to what recent projects like Athena have tried to do, you unlock a new asset class actually within crypto derivatives that can now be unlocked and traded. We're really interested in figuring out more creative ways to work with tokenizing different parts of the exchange and different parts of on-chain yield. That's super exciting. I hope we can see all of these things updated very soon.

I think that sets the conclusion of this pod. Thank you both very much for joining. Thank you so much for having us.