Unlocking Data Science for web3

Why I'm joining Variant as their first Data Scientist

I’ve joined Variant as the firm’s first Data Scientist. I could not be more excited to work with Spencer Noon, Li Jin, Jesse Walden, and the rest of the team to support and invest in the next generation of world-changing web3 companies. 

Over the last few years, data science and analytics have emerged as one of the most useful tools for understanding blockchains and the applications built on top of them. While data from web2 applications is private and only selectively disclosed, data from web3 applications streams in real-time and is publicly available to anyone in the world: you have the entire universe of user activity at your fingertips, and you can use that data to better design your product or make better informed investment decisions. I’ve joined Variant to do just this in support of both our world-class founders and our investment team.

How data science empowers founders

At the most basic level, data science helps teams better understand who their users are and what to build for them. It allows them to observe the activities of their users on-chain, understand if their product resonates with them, and improve their user experience. It’s my goal at Variant to help our portfolio companies level up their understanding using data analysis so they can build the best products possible, and ultimately find product-market fit.

To illustrate how this might look, let’s examine Sudoswap, a new decentralized NFT marketplace protocol that allows users to create liquidity pools of NFTs, and specifically how it was engaging users from multiple ecosystems. 

The first graph above shows the total ETH volume on Sudoswap compared to the volume that came from NFT-native users, or users who were already active in the NFT space on other marketplaces. When we compare these two metrics together, we see how Sudoswap was not only successful in onboarding the existing NFT community to its platform, but also an appreciable amount of net-new NFT users. Roughly 30% of their user base had never traded a NFT before! In the second chart I’ve highlighted this difference in volume between the two user groups and we can see that these net-new users were contributing up to 50% of volume within a single day.

Digging further into the data, we are able to see that many of these net-new NFT users were previously engaged exclusively within the DEFI market , an early but important indicator that Sudoswap is expanding the NFT market beyond its current set of participants. This type of analysis could inform Sudoswap’s product roadmap – more specifically, whether it’s worth doubling down on building certain kinds of financial tools, or identifying alternative areas for investment, or maybe even doing both.

…and opens up new windows for investors

A key takeaway from my early conversations with Spencer about joining the Variant team was that as an early-stage web3 VC, we can leverage data science across so many areas, from sourcing and diligence to portfolio support and beyond. Data science is especially important when evaluating investments: in the vast sea of web3 narratives it can be easy to get distracted by the new, shiny thing. It’s only with a deep, data-focused understanding of on-chain indicators that we can ultimately gain a true picture of network health. (This was also the original goal behind OurNetwork.) Variant will leverage data science to see which startups have strong fundamentals backed by on-chain data, not just an appealing narrative. 

To highlight an example of on-chain indicators flashing a signal before price appreciation occurs I’ve included a historical graph below of Polygon. Polygon is a decentralized Ethereum scaling platform which has gained substantial adoption over its lifetime. In the graph below I have plotted both the price of the MATIC token on the right axis and the number of daily transactions occurring on the protocol on the left axis. The number of transactions occurring on the network itself is an important measurement that we can use to gauge adoption not just among users, but also developers building applications in its ecosystem. Notably, before MATIC exceeded $0.10, the network’s strong growth in transactions, which at times exceeded 100K in a single day, proved to be an early indicator of future ecosystem growth and price appreciation

From Mining, to Matcha

I’m coming to Variant from 0x, the DeFi exchange protocol that has done over $100 billion dollars in trading volume. There I led all data science on their end user product, Matcha, providing key insights from both protocol and product that drove our strategy. 

My journey to 0x and now Variant is just the latest chapter in a love affair with crypto that began a decade ago. I first started mining Bitcoin during graduate school, in 2012. My career initially took me to the mobile gaming industry, where I built in-game economies and analyzed user behaviors. Then, at a media startup, I studied the relationship between the geographic proximity of a news story to viewers and the degree to which they valued it. What tied my work together was an enduring interest in using data science to level up customer experiences and understanding. As web3 goes mainstream, there’s a clear need for this same work.

Building the user-owned web

After meeting and getting to know the team, I was moved by Variant’s deep passion for building the user-owned web and supporting their world class founders. This was critical to me in transitioning from working on product teams and choosing my next role. I didn’t just want to invest – I wanted to help build the future of web3. 

The magic of data science is that it tells us not what we want to know, or think we know, but what we need to know. In web3, our collective data set is so much more open and democratized than anything that has come before it. I cannot wait to start applying my skills to unlock insights from these data sets in the months to come.  

If you’re interested in web3 data science, or investing – let’s talk. My inbox is open: [email protected]