How to Do a Crypto Airdrop That Actually Works
Airdrops -- sending free tokens to users’ blockchain addresses -- can be a powerful crypto marketing tool. Done right, they can build brand awareness, foster community, and kick-start a new token economy.
In 2017, a blockchain startup called OmiseGO launched a batch of airdrops. Hoping to spread their token widely and give something back to the Ethereum community, OmiseGo decided to airdrop tokens to ETH holders with an account balance of above 0.1 ETH, even if those people hadn’t requested or even heard of OmiseGO tokens. Since then, many token projects have replicated the OmiseGO model.
But does this popular approach actually work? Many people who receive an airdrop invitation treat it as spam. Others may accept it but still have no interest in using the tokens they receive. Giving away tokens to users who ignore them isn’t an effective community-building strategy.
How can we make airdrops more efficient? We analyzed the data to find out.
To showcase our analysis, we’re using a hypothetical airdrop on the Stellar network as an example, but this analytical approach would work on other blockchain networks as well.
So let’s pretend we’ve launched a new token on the Stellar network, and we want to do some airdrops for promotional purposes. We don’t want to waste money sending tokens to dead accounts or to people who’ll never claim or use them. How can we identify the most ideal recipients for our tokens?
First, it would make sense to look at the last month of transaction data from the Stellar network and identify active users as targets for our airdrops. We might also want to see how many tokens those users already have, on the theory that users with more tokens in general may be more likely to consider and use a new token. Ideally, we want to airdrop tokens only to the people most likely to actually use them.
So what does the data look like? First, let’s analyze the balance distribution of Stellar’s XLM coins at each address on the Stellar network.
This chart shows us that 87% of the addresses on the Stellar network have a balance of less than 10 XLM. This is a useful data point for a new project, but it doesn’t offer us a lot of guidance.
Looking at a user’s historical trading behavior would probably be a better indicator. How many coins a user has isn’t necessarily the most important data point. What we want are users who are willing to do something with the coins they have. Active users are likely to be better for brand communication and network growth.
If we count the number of transactions for each account in the Stellar network over the last month, we find that more than 89% of the accounts had fewer than 10 transactions. Furthermore, we found no clear connection between the balance of each account and its number of transactions. Large balances do not necessarily mean a large number of transactions, and small balances do not necessarily mean low transaction frequency.
What we want, ideally, are users with high balances and a high transaction frequency. To find those users, we can use the following simple formula to assign a unique score to each account.
score = log(balance * number of tx)
This approach gives us a single score for each user that factors in both transaction count and account balance. (We take the logarithm here to keep the final score numbers from being large and unwieldy). The higher a user’s score, the more likely they are to be a good target for our airdrop.
The following table shows the top 20 account addresses calculated in this way:
This method of analysis allows us to filter out a large number of invalid and zombie addresses, and focus our airdrops on users that are both invested in crypto and actively involved.
If we want to go deeper, more advanced data analysis methods will allow us to study users’ holding or spending patterns, which would be an even better indicator than just looking at account balances and transaction counts. Ideally, we want to analyze the length of time an account holds a coin, in order to distinguish between long-term and short term investors. We could also analyze users’ past transactions to see if they have any history of dealing with shady accounts.
Doing this kind of analysis will allow blockchain projects to better understand the targets of their airdrops, and avoid wasting tokens on inactive accounts.