Why product analytics in web3 is a challenge
As a product manager of DeFi product that has entered a growth stage, I saw my team stumbled upon a question one day: “Where should we scale next?”. Aiming to find non-random answer to this question, I spent the last few months focusing on setting up a data analytics process to understand our users, their segments, profiles, and behavioral patterns.
Thanks to powerful product analytics tools like Segment, Amplitude, Mixpanel, GA, modern product teams can easily track user actions, conversion rates, predict churn and segment users. However, as soon as I started trying to put our data together, a lot of new unique challenges came up. When you’re building a Web3 product, the name of the main of them is — Decentralization.
Decentralization is beautiful, but challenging, as it fragments user identities across multiple wallets and chains, making it difficult to get an accurate picture of users’ profiles and segments. When it comes to data, in web3 it’s relatively easy to collect and analyze everything about transactions, volumes, tokens, wallets, thanks to the public nature of on-chain transactional data. But behind every wallet address and transaction stands a user. When you buy an NFT on Opensea or swap tokens on Uniswap, there’s a lot of context leading to this action. And it’s extremely hard to link the on-chain behaviour to this context.
This occurs due to several principles of how web3 works:
- No KYC, no identity. Decentralized apps don’t require any form of registration or KYC and don’t store “user account” that can be associated with user’s in-app behavior.
- Decentralized wallet, decentralized identity. Wallet addresses contain only transactional data, which is publicly available on the blockchain. This design keeps on-chain behavior separate from the identity.
- Multiple wallets, multiple identities. Users tend to have several wallets for different use cases — for example, one to hold tokens and one for trading — which allow them to multiply their identity (and skew numbers) indefinitely.
The truth is, not all engagement data comes from the blockchain. Analyzing only on-chain data is not enough for making decisions and building a complete picture of users. They still use Web2 interfaces in between interactions with the blockchains, so missing all the stuff users do on Web2 is like staying blind.
Here’s what I mean.
Identity and interface fragmentation makes it hard to match actions done on Web2 UIs (eg. clicks, page visits, etc.) with on-chain ones (eg. transactions, swaps, transfers, etc.), and trace them both back to one, single identity (wallet). Moreover, if trying to solve this, it’s extremely important to preserve the privacy of users, based on the web3 principle of data ownership.
So, all the web2 analytics tools work not so well with web3 products. I believe that new, hybrid data analytics solutions will be required to unify data fragmented across Web3, trace it back to Web2, have an accurate picture of users’ profiles and segment them properly — all while preserving their privacy.
I’m really excited to see, what these new tools will look like and what they will be able to offer to the web3 product teams striving for data-driven growth? What are your thoughts? Please, share.
