Crypto Growth Hacking w/ on-chain data
How we found 310,888 potential Axelar GMP Power Users to target
It’s hard enough to get users on 1 blockchain. Try 10.
Axelar is a cross-chain transfer and messaging protocol focused on EVM <-> Cosmos transfers of tokens like USDC (and synthetic axlUSDC), ETH, AVAX, etc.
As one of our partners we’ve been knee deep in their data figuring out the best way to connect what’s happening across different chains with the goal of helping them grow. So we did an EVM analysis to identify power users, take those profiles, and identify market opportunity among active addresses.
Starting Small w/ Squid
On the data science side, we focused on transfers via General Message Passing (GMP), most specifically the Squid Router.
Squid lets people take a token on 1 chain (ETH on Ethereum) and turn it into a different token on another chain (MATIC on Polygon). It coordinates DEX swaps as needed and uses the GMP transfer of the main bridge token (typically axlUSDC which it can burn & re-mint) as the intermediary.
We went to each of: Arbitrum, Avalanche, Binance Smart Chain, Ethereum Mainnet, and Polygon to identify sent messages (most often USDC or axlUSDC) to other chains including each other but also Osmosis, Fantom, Moonbeam, etc.
Transactions originating from Ethereum Virtual Machine (EVM) blockchains use the chain’s local Squid Router address:
with squid_to_burn AS ( SELECT BLOCK_NUMBER, BLOCK_TIMESTAMP, TX_HASH, ORIGIN_FROM_ADDRESS as EOA, CONTRACT_ADDRESS as token_address, event_inputs:value as raw_amount FROM -- swap out each chain's event logs bsc.core.fact_event_logs WHERE EVENT_NAME = 'Transfer' -- Squid Router burns token AND EVENT_INPUTS:from = '0xce16f69375520ab01377ce7b88f5ba8c48f8d666' AND EVENT_INPUTS:to = '0x0000000000000000000000000000000000000000' ), ...
For those who like to jump straight to SQL, here’s the Flipside query for BSC Squid transfers.
Chain specific table showing which Externally Owned Account (EOA) called Squid, from what chain, to which chain, for what token, in what amount.
We then explored network graphs to see which chains seemed to be most connected. This could help us hone in on specific clusters of chains for Axelar to focus its growth resources on.
Because Squid is still new, we parsed out the difference between active users who are OG (have been using Squid since launch) vs New and Active (started in February and use it regularly).
We then used our crosschain tags table:
crosschain.core.address_tags to identify a few broad factors that we thought could explain why some people find a fit with Axelar over others:
Are addresses who use multiple NFT platforms (opensea, looksrare, x2y2, rarible) the kind of “experimenters” that dabble on different chains and need bridges like Axelar?
Are addresses who use multiple Central Exchanges (or have a “main” address known to a CEX) more likely to be re-used across EVM chains (where the addresses all match across the chains)?
Does actively using more chains increase the likelihood of trying out multiple bridges? Or do people just use the “official” bridge (e.g., Polygon Plasma) and keep their chains to a minimum?
We found that new active users that were “sticky” to GMP (using it 5+ times in their first month) were especially likely to:
Use 3+ chains (not just 2, but not 5 like the early users)
Actively traded NFTs compared to Persistent Early Users
More likely to interact with a central exchange at least once.
Identifying Addressable Markets
Of the 399,587 EVM Addresses that met the following criteria:
Active on 3+ EVM chains ever
At least 1 transaction from or to a central exchange (on ETH)
At least 1 transaction on an NFT Marketplace (on ETH)
310,888 were active on any chain in 2023.
Of those active in 2023, the chain popularity order is ethereum (required to meet criteria), polygon, bsc, optimism, arbitrum, avalanche.
Our Target Addressable Market Summary
This leads to directly usable feedback for Axelar:
Optimism is an EVM Layer 2 they don’t support, that co-occurs highly with potential users.
There are known clusters like Arbitrum ←→ Polygon; BSC ←→Avalanche.
Some protocols (e.g., AAVE) operate 3+ of these chains. Is there partnership potential?
With 310,888 addresses active in 2023, allowlists can be coordinated with protocols live on multiple clusters of EVM chains.
Axelar’s primary Cross Chain Transfer protocol, Satellite, has more transaction history in the EVM ←→ Cosmos market, but focusing in on Squid was useful as it seeks to grow Axelar’s market share in the popular EVM space with stiff competition like HOP (whose airdrop we’ve written about).
Lookout for new Axelar tables that make Squid and Satellite cross-chain analysis much easier in the Flipside data app. And for those who like code here’s the full report & reproducible repo.
The fully reproducible GitHub Repo