The airdrop trilemma: Why can’t we do a one-time airdrop?

Original title: "The Airdrop Trilemma: A Data Problem in Disguise."

Original author: KERMAN KOHLI

Original compilation: Luccy, BlockBeats

Editor's note:

In the crypto world, airdrops are both anticipated and controversial. Although people have become accustomed to this, problems still arise frequently. Some believe that insiders only want to cash out by selling their tokens, or that the team lacks understanding and suitable advisors. Others believe that whales should be given more priority because they bring more liquidity, while others believe that airdrops should embody the democratization of cryptocurrencies.

Crypto researcher KERMAN KOHLI explores the fundamental choice dilemma behind airdrops, pointing out the balance between capital efficiency, decentralization and retention. He analyzes Optimism’s four-round short metrics and data, highlights the importance of retention, and offers suggestions for segmenting airdrops and obtaining better data. BlockBeats compiles the original text as follows:

Recently, Starkware launched its long-awaited airdrop. Since most previous airdrops caused a lot of controversy, people gradually stopped being surprised by airdrops, and the same is true this time.

So why does this happen over and over again? Here are some of the opinions one might hear:

· Insiders just want to sell and cash out billions

· The team responded to this No clue and no proper advisors

· Whales should be given more priority because they bring TVL

· Airdrops are about democratizing what cryptocurrencies are about

· Without farmers, there is no use or stress testing of the protocol

· Misaligned airdrop incentives continue to have strange side effects

None of these views are wrong, but none of them are entirely correct in themselves. Let’s unpack some of the key points to ensure we have a comprehensive understanding of the issue at hand.

There is a fundamental dilemma of choice when doing an airdrop, you need to choose between three factors:

· Capital efficiency

· Decentralization

· Retention rate

You will often encounter situations like this: Airdrops work well in 1 dimension, but rarely strike a good balance between 2 dimensions or all 3 dimensions. Retention in particular is the hardest dimension, anything over 15% is usually unheard of.

· Capital efficiency is the criterion used for how many tokens are provided to participants. The more efficiently you distribute your airdrop, the more it becomes liquidity mining (one token for every dollar deposited), benefiting whales.

· Decentralization is defined as who gets your tokens under what criteria. Recent airdrops have adopted an arbitrary standard approach to maximize the number of users who receive said tokens. This is usually a good thing, as it saves you from legal trouble and gives you more leverage to make people rich (or pay their parking fines).

· Retention rate is defined as how much users stay after the airdrop. In a sense, it's a way to measure the user's alignment with intent. The lower the retention rate, the less consistent users are. As an industry benchmark, a 10% retention rate means that only 1 in 10 addresses are actually correct.

Retention rates aside, let’s examine the first two in more detail: capital efficiency and decentralization.

To understand the first point of capital efficiency, let us introduce a new term called "Witch Coefficient". It basically calculates how much you'd benefit from splitting a dollar of your money across a certain number of accounts.

You are on this spectrum The position will ultimately depend on how wasteful your airdrops will become. If your witch coefficient is 1, technically this means you are running a liquidity mining scheme and will piss off a lot of users.

However, when you get to something like Celestia where the witch coefficient balloons to 143, this is extremely wasteful.

Decentralization

This leads to the second point about decentralization: You ultimately want to help the "little people" who are real users, even if they aren't rich, and seize the opportunity to use your product early. If your witch coefficient is too close to 1, then you give almost no help to the "little guys" and the maximum profit goes to the "whales".

Now, this is where the airdrop debate gets heated. There are three types of users here:

·The "little guys" are here to make a quick buck and move on (probably using some wallets in the process)

·"Little guys", they are here to stay and like the products you make

·"Industrial farmers, they act like a bunch of little guys", they are definitely here to accept your Most of the incentives and sell them before moving on to the next thing

The last one is the worst, the first is still acceptable, the second is the best. How we distinguish these three is the biggest challenge of the airdrop problem.

So how to solve this problem? While I don't have a concrete solution, I do have a philosophical approach to how to approach this problem that I've been thinking about and seeing firsthand over the past few years, which is project relative segmentation.

I'll explain what I mean. Zoom out and think about the meta-problem: you have all your users, and you need to be able to segment them into groups based on some value judgment. The value here is observer context specific and therefore will vary from project to project. Trying to attribute some "magic airdrop filter" is simply not enough. By exploring the data, you can start to understand what’s really going on with your users and start making decisions based on data science on the appropriate way to execute airdrops through segmentation.

Why doesn’t anyone do this? This is another article I will write in the future, but it is long and about hard data problems that require data expertise, time, and money. Not many teams are willing or able to do this.

Retention

The last dimension I want to talk about is retention rate. Before we talk about it, it’s best to first define what retention means. I summarize it in these points:

The classic mistake most airdrops make is treating it as a one-shot equation.

To prove this, I thought some data might be helpful. Optimism has actually performed multiple rounds of airdrops. I wish I could find some simple Dune dashboard that would give me the retention numbers I want, but there isn't one. So, I decided to roll up my sleeves and get the data myself.

Without overcomplicating things, I want to understand a simple thing: how the percentage of OP users with non-zero balances changes over successive airdrops .

I go to This website obtained a list of all addresses participating in the Optimism airdrop. I then built a little scraper that manually fetched the OP balance for each address on the list (burning some internal RPC points for this) and did some data wrangling.

Before we dive in, it's important to note that each OP airdrop is independent of the previous airdrop. There is no bonus or link to the previous airdrop tokens. I know why, but let's continue anyway.

Airdrop 1

Granted 248,699 recipients, The criteria are as follows :

· OP Mainnet Users (92k addresses)

· Duplicate OP Mainnet Users (19k addresses)

· DAO Voters (84k addresses)

· Multisigners (19.5k addresses)

· Gitcoin Donors on L1 (24k addresses)

· Ethereum Pricing Users (74k addresses(), 0

After analyzing all of these users and their OP balances, I got the following distribution. Balances represent users dumping unclaimed OP tokens after they were sent directly to eligible addresses at the end of the airdrop (according to dune data).

Anyway, this first airdrop was surprisingly good compared to previously executed airdrops I’ve observed. Most had a>90% dump rate. Surprisingly, only 40% of people had a 0 balance.

I then wanted to understand how each criterion performed in determining whether a user was likely to retain their tokens. The only problem with this approach is that addresses can fall into multiple categories, which skews the data. I won’t take it at face value, but here’s a rough metric:

I then wanted to understand how each criterion performed in determining whether a user was likely to retain their tokens. The only problem with this approach is that addresses can fall into multiple categories, which skews the data. I’m not going to take it at face value, but here’s a rough metric:

Close to 90% of addresses holding 0 OP balances is the usual airdrop retention stats one is used to seeing. I’m tempted to dig deeper into this, but I’m keen to move on to the remaining airdrops.

Airdrop 3

This was by far the best executed airdrop by the OP team. The criteria is more complex than before, and has an element of “linearization” mentioned in previous posts. This was distributed to around 31k addresses, so smaller but more efficient. The details are summarized as follows:

· OP Delegated x Days = the cumulative sum of OPs delegated per day (i.e. 20 delegated OPs for 100 days: 20 * 100 = 2,000 delegated OPs x days).

· Delegates must vote on-chain in OP governance during the snapshot period (0:00 UTC on January 20, 2023 and 0:00 UTC on July 20, 2023).

A key detail to note here is that the criteria for on-chain voting is a period of time after the last airdrop. So the farmers who come in the first round think "Okay, I'm done farming, it's time to move on to the next thing". This is awesome and helps with analytics because look at these retention statistics.

Among these airdrop recipients Only 22% of the tokens have a balance of 0, which to me indicates that this airdrop was far less wasted than any previous one. This fits in with my argument that retention is crucial and having the extra data of multiple airdrops is more useful than people give it credit for.

Airdrop 4

This airdrop gave a total of 23k addresses, and there is an update Interesting standards. I personally think this retention rate will be high, but after thinking about it, I have an argument to explain why it might be lower than expected:

· You are on the hyperlink Engaging NFTs created. The total gas in transactions on the OP chain (OP Mainnet, Base, Zora) involving the transfer of NFTs created by your address. Measured 365 days before airdrop deadline (January 10, 2023 to January 10, 2024).

· You created a compelling NFT on the Ethereum mainnet. The total gas in transactions involving transfers of NFTs created by your address on Ethereum L1. Measured 365 days before airdrop deadline (January 10, 2023 to January 10, 2024).

Surely you would think that people creating NFT contracts would be a good indicator? Sadly not. The data suggests otherwise.

Although it is not like the The second airdrop was so bad, but compared to the third airdrop, retention has taken a big step back.

My hypothesis is that if they had additional filtering on NFT contracts that were marked as spam or had some form of "legitimacy", these numbers would show To improve. This standard is too broad. Additionally, since the tokens are airdropped directly to these addresses (without having to be claimed), you end up falling into a scam where the NFT creator thinks "Wow, free money, time to dump."

Conclusion

As I wrote this article and obtained the data myself, I managed to prove or disprove some of my hypotheses, which proved to be very valuable. In particular, the quality of your airdrops is directly related to how good your filtering criteria are. Anyone trying to create a universal "airdrop score" or use advanced machine learning models will fail, prone to inaccurate data or a large number of false positives. Machine learning is great until you try to understand how it arrives at its answers.

While scripting and coding this article, I was given numbers for Starkware airdrops, which was also a fun intellectual exercise. I will write about this in my next post. Key takeaways teams should learn from here:

· Stop doing one-time airdrops. You are shooting yourself in the foot. You want to deploy incentives, similar to a/b testing. Iterate a lot and use past experiences to guide your future goals.

· With standards built on past airdrops, you will improve your efficiency. Effectively, more coins are given to people who hold them on the same wallet. Make it clear to your users that they should stick with one wallet and only change wallets when absolutely necessary.

· Get better data to ensure smarter, higher-quality segmentation. Bad data equals bad results. As we saw in the article above, the less “predictable” the standards are, the better the retention results will be.

「 Original link "