Chainalysis has been been playing a key role in the crypto ecosystem. As the leader in the blockchain data and intelligence market, it’s made it easier for many financial and government institutions to feel more comfortable with the space.
Chainalysis has been growing fast, having raised $360 million of venture capital to date, including most recently in Series E at $4.2 billion valuation. Before founding Chainalysis, he was the COO of Payward, the leading Euro to Bitcoin exchange. Importantly he was also a co-founder of the crypto exchange, Kraken, in San Francisco. He holds a PhD in quantum mechanics, no big deal.
Below is the video and full transcript.
(Crypto Driven is a team effort – many thanks to my FirstMark colleagues Jack Cohen, Karissa Domondon and Diego Guttierez)
VIDEO:
TRANSCRIPT [edited for clarity and brevity]
[Matt Turck[ What is Chainalysis and what is the origin story of it?
[Michael Gronager] Before Chainalysis, I was at Kraken, the crypto exchange in San Francisco, and saw some parts of the early industry. I’ve been in crypto for, as you say, more than 10 years. And in the early days of the crypto industry, we were all curious about how to build a great company in crypto, how to be famous, how to do the right things and so on. And to be honest, the only thing you could really build in the early days, in the crypto space was exchanges because you needed people to get access to the crypto space. Very much analogous with the internet, initially, the only thing you could do was provide people access and you could build some content.
But basically you wanted access to crypto. So building the rails from a bank or from a credit card process and other things to buy crypto and get exposure to this new magic internet money, that was basically what you wanted to do.
Then we also heard a lot of rumors about crypto being used for criminal activity, people could buy drugs. There were rumors about terrorist financing. There were a lot of other things happening in the early days. And I got quite involved in the Mt. Gox hack and looked at that big exchange hack. It was roughly half a billion dollars that had been stolen there. And there was a huge criminal activity there. But the problem around us, no one could really figure out what happened.
And when I spoke to law enforcement in Japan back then, they were basically like, “Well, we don’t really know. It went into crypto and now it’s gone.” And that was the perception of the crypto space, that this was anonymous and that was probably how it should be. So at the same time, working further on at Kraken and looking at like, how could we set up shop in Japan? How could other things happen? Was an important part.
Talking to regulators in Japan, talking to banks all over the world, I kept hearing the same concerns. How do you do transaction monitoring in crypto? How can we even understand the origin of funds? How do we even assess how much is being used for various purposes? And how can you even get some transparency into this cryptographic value transfer system?
I suddenly realized that the answer to this was hidden in plain sight. All of the transactions on a blockchain are freely available. They are on the blockchain, all the billions of transactions, but the challenge here was to understand how to interpret them. What was actually going on in the blockchain?
And then I said like, “Okay, can you build that algorithm that can somehow give an image, or picture, or map of the blockchain that will enable people to understand what happens here and enable you to build all kind of data you want on top of that?”. That was really the core challenge that I tried to solve initially.
The way I usually explain what this is all about, you have two transactions on a blockchain and you want to assess whether they’ve been done by one individual. That’s basically the same party that transacts there or two different instances. Then you can run that algorithm over billions of transactions. On the other end, you’ll get millions of different wallets that’s been transacting on a blockchain. Then you can start looking at the individual wallets and say, “Oh, this is only transacting in $5 at a time. So it’s probably like a micropayment service.” Or, “This one is transacting in bigger amounts, so it might be an exchange. We can see when it’s active. It’s only active in daytime hours in Asia. And so are all the wallets interacting with that. So it’s probably an exchange in Asia.”
So suddenly you can create a lot of understanding of what is crypto actually about from a real world entity thing. So fast forward, we then build the biggest database in the world of connections between real world entities, being exchanges, dark net markets, everything that you can find in the crypto space and digital identifiers on the blockchain. So you have a transaction and immediately you know who sent it or received it. That is the basis foundation for everything we build following that, and everything on top of that is then the products. And it’s like how do you monetize data? You do it through products and products can be investigation, compliance, and so on.
Do you cover all blockchains? Does L1 versus L2 make any difference? How do you think about it?
We cover everything. The only right answer to this is basically everything. It doesn’t mean that everything is there as of today, but if it’s not, it’s there tomorrow, or it’s in the pipe of happening. Of course, there’s a little bit of lead time between the genesis block of a new token project and when we have it in our product. But in principle, everything is supported from the onset and that’s the focus there. We launch support for Lightning, one of the Layer 2 projects that’s there. We have other Layer 2s coming. So it’s like for me, it’s everything that moves on public blockchains is something that we have support for in the product.
You mentioned having a database where you keep a record of everything, it must be absolutely gigantic. How do you store, and manage, and query a database of that size?
The best way to think about it is that, you have all of the transactions on the blockchain. Well, they’re kind of storable because by definition, someone has to store them, otherwise you can’t really build a proper node in a blockchain network. So all of those are storable. We keep them stored, like have a lot of updates around that. And then for each single transaction, we store metadata. What have you figured out about those transactions? One algorithm might solve that that was sent from Matt to Lisa. Another algorithm might say that, “No, this was a treasury transaction inside a big exchange.” And then all of that information is available and we continuously build that on top of that.
So think about that as one big spreadsheet, where it says, “Transaction from blockchain sent maybe by to maybe to.” And then we build out a big database of that. And then on top of that, we can now run different queries. We can aggregate to say, what is the biggest exchange? How many transactions did it send? How much did it send last week? Are people moving funds into that exchange or moving a fund out of that exchange? Do we see, for example, as in the beginning of COVID, we saw a lot of, I would say smaller investors trying to get liquidity by selling their Bitcoins. And that was the price that went down. And there was some panic around that. But actually, what happened was that the big whales were just buying and the small investors were selling. And that, again, you can say a lot about what happens in the market from data like that, for example.
Presumably you have a lot of machine learning to identify patterns?
Yes, we do. Again, it’s like, if you build a simple algorithm, are these two transactions done by the same or others? While you can create something and then suddenly you get a signal and you have something that looks about right. But of course it’s not precise. And you can take the analogy to the paid track algorithm that Google invented initially. Of course, it kind of worked initially and you were all happy about it because it was better than AltaVista. But it was not really anything that was amazing. Then over the years they started to improve that with a lot of manual search optimization.
Then in the end, a lot of machine learning algorithms hadn’t… Suddenly there was no ground truth data to enable you to run machine learning algorithms and get them adjusted and optimized and suddenly you get good quality. We’ve done the same over the years. So basically in the early days, we had simple algorithms, and improved that over the years by getting access to ground truth data. Then once we use that, we could optimize the algorithm. Then when we look at the data today and run the same algorithm and the data, it’s like several hundred times better, because you just have all the training data, worked with the customers and law enforcement to get access to that. Suddenly we have a very good understanding of, I would say what’s happening on a blockchain.
I’d love to go into those various products that you built. You had this initial technology and the database, and then you very smartly built this into multiple products, serving multiple needs. From my notes here, I see there’s Chainalysis Business Data, there’s KYT, there’s Kryptos, Market Intel and Reactor. Can you give us a little bit of a whirlwind tour through the various products and what they do?
It all started with Reactor. So once I had the first database and a backend that kind of worked, you could run different queries. But you soon realize that if you try to sell that to anyone, they’re like, “I don’t know what it looks like. I need some UI in front of that.” So building Reactor was the first input here. So basically visualizing the map of the blockchain. So Reactor is the way that we sell it today. And what I would call it is like an investigation or enhanced diligence product, where you, for example, can look up a transaction and immediately you would see this is the entire wallet from that service.
Then you can follow another wallet and say, how did they interact with each other? And there might be other wallets that you want to understand, how are they interacting? You might have a lead in a case, if you’re talking law enforcement, there might be a specific lead and they want to follow the flow of funds until they find, “Ah, that’s actually early days. The same person, our target here used a crypto exchange in, let’s say Slovenia.” And we can reach out to them and ask what’s the identifying information of the person and suddenly figure something out. So that would be an example of a use case for Reactor. It’s also being used for enhanced due diligence in a compliance setting and just for people to click around and trying to understand what is crypto actually about. So that’s Reactor.
Then in around 2018, we launched KYT. And KYT was launched as a reaction, seeing civil crypto exchanges having hundreds of accounts of Reactor. They were basically going through all transactions above a certain amount and tried to understand what happened. But that doesn’t really scale when you move to millions of transactions. You can’t investigate all of those. So we created KYT as a transaction monitoring capability where all transactions are screened, and then only those who show any kind of, I would say, suspicious behavior, either in structuring, or it could be interaction with dark net markets, or illicit sources, other things. Then we would surface that through the compliance person and they would run it in Reactor and figure out what actually happened. So that was the KYT piece.
The other thing we saw was a need mainly from the established financing institutions. If they wanted to bank a crypto exchange or other crypto service business, I kind of asked at a high level, what are they doing? And then we said, well, if we aggregate a lot of data and try to build cohorts around it and so on, then they can say, “Oh, this customer that approached us now, they’re actually the second biggest exchange in the Netherlands. They have customers typically getting amounts of $200. That’s the amount they’re selling and trading and they’re using them, by the way, to do online gambling or maybe a gaming service somewhere.” And suddenly they got a good, I would say, helicopter view of the crypto space that made it comfortable for them to provide banking services to the different players in the space and also just understand what businesses in crypto were doing.
So that was Kryptos. Then moving into the business intelligence piece and the market intelligence stuff for that one. Market Intel is very much around some of the stuff I mentioned before. You want to assess what is going on in the crypto market today. Are people receive, the price goes up or we see it go down. What is the reason behind it? And if you look in the tea leaves of the blockchain, you’ll, for example, find examples that, oh, the small investors are trying to sell and liquidate the assets because they need liquidity right now. So we’ll see a lot of small transactions moving into crypto exchanges. But we don’t see any big transactions going into crypto exchanges, so we don’t see any whales leaving the space.
So that means that we expect that the market movement will be temporary. And there’s a lot of similar other things that you can assess from that. You can also say, “Oh, China is liquidating all the assets, but it doesn’t happen in Europe. So we know that the price moving comes from China,” for example. So these things, we can assess from the data and that’s of course, of high interest from regulators, of high interest from investors into the space to understand what is going on. And then the final piece is business intelligence, where using the same data just to understand what are people doing on your platform? For example, are they also playing other games than the game that you’re providing? If you’re looking at that space, are they buying other NFTs? What is going on there? That’s business intelligence. That’s again like using the same data for another use case in a company.
Can you talk about some fun examples or fun stories? I know that there’s only so much you can talk about and some uses of the platform are very confidential. But like anything you can talk about, like fun stories.
Yes, I think there’s a couple of those. I think it’s always tempting to grab some of the law enforcement stories because they always do see and people always like to hear about it. But say, for example, we had the colonial pipeline hack. We released a blog around that one, how we collaborate with the API around API, used our products to basically figure out what went up in that case. Without going in too much detail around what actually happened there – what you can do is that you can identify, for example, a ransomware payment. This case, it was ransomware payment. You see what wallet it goes to, you see how it’s being moved around on the blockchain after that. And you suddenly realize that that movement is part of a bigger wallet.
And now you know, okay, that wallet has been active before. And then you realize what exchanges have they sent test transactions to, where are they intending to, for example, liquidate the funds? And you realize that. Then they might go through a mixer. They might do other things and obfuscate the flows. But you also already have identified because of earlier transactions, that they might use a certain exchange to liquidate the funds. Now, if you’re law enforcement, you can reach out to that exchange and say, “Hey, if you get a payment there, maybe you should freeze the funds and then we can get them back.” So that happened successfully there. I have another fun story that’s way closer to home. This happened, I think it was between Christmas and New Year. My girlfriend comes to me, is like, “Hey, my Instagram account has been hacked.”
And then someone reaches out on WhatsApp. She gave me her phone and then we start looking into what goes on. So someone is asking for ransomware payment for giving her account information back. We reached out to Instagram and have to run this parallel track between getting the account back by using the support system of Instagram, which might not always be the easiest in the world. At the same time, trying to ensure that the person is not deleting the account or anything, so doing that. In that situation, I started chatting with him. He wants to get paid $1,000. I negotiated down to 700. I have found his email address being used in other hacks and seen what happened there, realized he wants Bitcoin. So then I asked him, “Well, how do I know that I get the account back?” And then he actually sends me 10 other accounts that he had hacked and their statements that they got the account back, including Bitcoin transactions.
Reference accounts.
Reference accounts, yeah. So I ran this through Reactor and realized what exchanges that he’s using. I understand where in the world he is, can reach out to these exchanges that are customers of Chainalysis and ensure that these accounts get frozen. And then when I pay him $700, then I already know that I’m sending them to an account that’s frozen already. We got the account back from Instagram the next day, so it was not really a bigger deal than that. And I think the challenge here is I could see on the blockchain that he probably stole or extorted around a million dollars in three months from various people.
But it’s very, very hard to find the law enforcement agency that wants to run that case because for them it’s $700. And I can’t approach anyone and say, “I got $700 stolen in Bitcoin.” And they’re like, “Okay, what? That’s not interesting.” So that’s just another example. What we are trying to do in these cases is of course to find, okay, present the entire size of the case and go to a local law enforcement. And sell it to them and say, “This is actually something that could be solved. And this could make life much better for a lot of people, if you went after this case.” So that would be another example of stuff happening on the blockchain.
Chainalysis has been a gateway for a very broad group of financial institutions and traditional actors to get into the space, by providing some comfort that they will understand what’s going on, the data. And it will feel a lot more transparent. Where do you think we are in terms of traditional financial institutions getting into the space? What are you seeing?
We’ve seen over the last two or three years, some of the payment processing and neobanks becoming very interested in crypto. We’ve seen that from Square (Block), so they obviously become very interested in crypto data. And we saw it from Robinhood, we see it from PayPal. They want to kind of do something in the crypto space. It’s kind of typical. You see the newer player in finance… They grew a lot from just doing finance on the internet in the early days. So they clearly see that this new technology and that can properly grow their wallet share amongst their customers even further. So they entered the space and had a lot of appetite for that.
Today, it’s become like a core part of the revenue for many of them and clearly something that’s important. So that’s like today, all practical purposes, you would call them crypto businesses. Then what happened a little bit more than a year ago were that some of the more traditional financial institution, they saw the same. They saw that the price of crypto were growing, the adoption were growing. A lot of the high net individuals that were using their private banking service, would ask them, “How do I store my $10 million in Ethereum, ideally want to store them in your bank, but you don’t provide custodian services in crypto.” Then we saw an influx of enabling custody on traditional financial institutions.
Some financial institution started to offer buy, sell, hold of crypto. For example, Commonwealth Bank of Australia did that. That’s kind of the first step in this adoption phase into say, well, we are on the journey to become a normal crypto service business, now I called it normal. But basically do the same as everyone in the crypto space have been doing all the time. So I would say we are where I would say a lot of financial institution have become crypto businesses and most of them have a plan to become it. So that’s where beyond the adoption curve there. So I would say it’s getting there and it’s pretty close.
What are your thoughts on NFT, gaming, all those spaces? Are they relevant to you? Is that something that you think about?
Yes. I think the way look at this is, and others have made this analogy before, but comparing the growth of the internet with what happens in the crypto space. In the early days on the internet, there was not really much to look at. And then at some point, a lot of content side started to appear, web2 started to become a thing. And suddenly we got like the real growth of the internet happening. I think that when you look at NFTs, when you look at gaming platforms and other things, you actually have content on the blockchain for the first time. So that’s the first time you really see a use case that’s inherently tied to the digital world, to the blockchain is transfer of value of, I call it stuff on the blockchain.
Before that, it was basically investments into the underlying asset and were driven from speculative or investment purposes. And now we clearly see this trend that people want to buy an NFT, play a game where the stakes are actually real money or the feeling of real money and the feeling of real rewards if you are good at the game and we are seeing that happening. So I think that they are extremely important for showing the future of the blockchain and for us, it’s the same. It’s like being able to follow what happens on the blockchain and understand the growth of that industry. We’ve released NFT reports, we’ve done other things around this space, understand how that space is growing and figuring out how we legal data can help the growth of that space the best way. And ensure that customers on those services get the best possible journey for those companies.
I’d love maybe to finish for the last few minutes, tell us about what’s next for the company. So maybe starting with the product roadmap, to the extent that you can talk about it. What are some of the things that you’re building?
Yeah it dovetails to the question you had before around NFTs and others. I would say when we looked at the early days of crypto, everyone that were a company in the crypto space, were a financial institution. And now, as we said in the question before that, every financial institution is becoming a crypto company as well. But they’re all regulated, they all need compliance solutions and so on. But if you now look at the next level of growth in crypto, around NFTs, gaming, trading cards, and other things, they are not regulated businesses. They are basically just selling things and objects on the blockchain. I think some of the things that they start to need and understand is to understand how their customers are using their platform.
So one scenario, if we start in the more nefarious side, is for example, if you have a platform or game, you could create 10 accounts, 100 accounts, and suddenly you could collude. And because you don’t really have a concept of the single user in web3, suddenly they’ll be smurfing or other interactivity where someone is trying to cheat a game, just like an online program if have seen that. Another situation would be just trying to understand how you should treat a certain customer that enters a platform. Is it someone that played a lot of games before? Are they an old [inaudible]? Are they a new? How can you best follow them through what happens on the platform and basically help them that way? And I think that’s another area and that’s where we look at like the business intelligence side and understanding the customer space there, where I’m very focused on that this year. And that’s clearly something we see as being a core focus of what happens in Chainalysis this year.
For the company itself, what’s super interesting is that in a world full of crypto projects, and tokens, and that type of thing, so ultimately you’re a B2B, SaaS company, right? Or you’re like a data platform. So is the end game like you become a global, I mean, you probably are already a global company. You do an IPO. Is that sort of how you think about it?
The way I usually phrase that internally and think about it is that we are in the process of growing up. We are right now, basically trying to mature the company in various ways. We built a really great leadership team over the last year, have hired the lot of great execs there. We are maturing the company in various ways with the right processes, the right structures internally, and doing all of that. We are more than 600 people today and growing to more than 1,000 this year. So clearly in hyper growth mode. So my focus is basically trying to ensure that everything grows roughly in sync.
And we have roughly the right assessments of where we should grow and then basically execute on this opportunity we have, how can we monetize the data that we have in various ways? And how can we best build products for the crypto space to accelerate the growth of the crypto space? Which has always been the core focus of Chainalysis. And then where that leads us, yes, we’ve been global. I think the first customer I had was in the US, the second one was in South Korea, then India, and then Europe. So we’ve always been around the globe on the customer side, and today we have offices everywhere. So clearly been being global from the onset and continue to grow and mature the company. Then at some point there’s logical next steps in that journey. But that’s not something that we are obsessing about yet.
Taking your Chainalysis CEO hat off for a minute, and just like you, Michael, what do you think is super interesting in the crypto web3 space these days? Like an area, or a project, or a company, or you’re like, “Oh this is really cool.” Like almost like I wish if I was not doing this, I’d do that type thing.
I have a little bit of I think maybe a pet project or something that I really think is cool technology, and that’s a Helium Network. I think that’s an interesting network because it basically connects the real world with the digital world in a very nice way. You can do things as proof of location and other things that you can use. And if you look at the adoption from a lot of IOT projects and others, they start to depend on Helium and the things are being built on top of that. I think that’s the first time we’ve seen that in the crypto space, where suddenly having crypto as a core piece of the infrastructure for how we are building other products in the world is happening. And yes, it’s all in the early stages and all of that, but still, I think we see some very clear signs of how the world is changing. And I think that’s extremely exciting. So yeah, that’s one of my interests in the crypto space right now.
Okay, very cool. We happen to be very early investors there, but I promise this was absolutely not planned, so thank you for picking that one of all. Cool, thank you so much. Look, I’m super grateful and it’s such an interesting story. We had your co-founder, Jonathan, at one of our events, one of our Driven events back in 2018, and the company already had a nice velocity. But it’s really amazing what you all, like the 600 of you have achieved in the last few years. So I’m again grateful and looking forward to seeing the journey ahead. And maybe you can come back in a couple of years and as an even bigger company CEO and tell us the next part of the journey.
Would love to, thanks.