Yes, Airdrops are Dumb. But they don’t have to be. This reaction to this post really got me thinking. Here's a question: Why do IPOs always pop? Simple—it's by design. Every company wants holders instead of dumpers on their cap table. Institutional investors like BlackRock and Fidelity are the long-term holders that every CEO wants as shareholders, so they get offered shares at a discount to where the market is expected to clear. That discount creates the IPO "pop." Retail doesn't get that discount because retail is a swarm—some are holders, some are dumpers, and companies can't tell which is which at the IPO. So retail pays market price. The same dynamic plays out in crypto. VCs and institutions have legible long-term reputations that make them easier to differentiate from mercenary capital. The best value-add investors get preferential access, while retail pays sticker price. But airdrops happen on the transparent blockchains, where you can see which wallets are which. So teams use on-chain data to filter our “farmers” or sybils—people with thousands of accounts faking metrics just to get an airdrop. And yes, that makes sense. But nobody seems to be trying to figure out who is actually going to hold their token or dump it—who are the little baby Blackrocks and Fidelitys who deserve to be rewarded alongside the others value-add investors. Why don't projects do this? The Current State of Airdrops We all know airdrops are broken. Projects spend months attracting farmers who generate artificial activity, only to watch those same farmers dump tokens immediately after TGE. It seems the only solution people propose is to pivot away from airdrops to crowdsales. But it's 2025 now—there's a larger design space we haven't explored. Some projects have moved partway there. Optimism, Arbitrum, and Kaito have all modified their post-TGE incentives to reward long-term holders of their own tokens. But this strategy only works after your token exists. At initial distribution—usually the largest in dollar terms—you don't yet know whether users will hold or fold. The mistake these distributions make is trying to anticipate user behavior solely toward their own token. Instead, you should reward users based on how they've behaved with previous tokens. When BlackRock gets IPO allocations, companies don't know if BlackRock will dump their shares. But they know BlackRock generally hasn’t dumped previous IPOs. They value BlackRock on their track record, rather than by directly tying their hands. It’s crazy that token distributions don’t work this way. To fix airdrops, we need meta-incentives. Your airdrop should incorporate how users behaved in previous airdrops. Once users receive your token, you then need to make their behavior legible to the next project considering an airdrop. Here's a sketch of how this could work: After the airdrop, most teams just publish a list of allocations. Instead, they should continue to publish a Holder Score that updates after TGE: percentage retained over time, delegation/staking/voting participation, product usage, fee payments, liquidity provision, builder contributions. If users know future protocols will see this Holder Score and incorporate it into their own airdrops, those users will adjust their behavior today. This creates a meta-incentive—after your airdrop, you no longer have leverage over users, but the next project is implicitly collaborating with you to enforce that meta-incentive. The airdrop meta already did this once, making all projects attract farmers even when they weren't themselves planning airdrops. We can do it again and reward the best users through holder scores. When Airdrops Still Make Sense The strongest case for airdrops is pay-for-performance scenarios. If your protocol needs TVL, volume, open interest, or liquidity, you can incentivize that with points and convert linearly to tokens. This kind of airdrop will never go away because it directly offers rewards for measurable value. But then you have amorphous airdrops—for layer 1s, infrastructure, or consumer products, where it's unclear what metric you should be optimizing for. For these, we can do better than airdrops. Of course, it’s fine to airdrop small amounts to targeted groups: direct contributors, power users, early supporters, or adjacent communities. But broad helicopter money airdrops just don’t work—they only incentivize farmers to generate artificial activity that disappears after TGE. That's useless for everyone, including founders and other tokenholders. Instead of airdrops, let early users earn the right to invest at preferential prices in the crowdsale. Once you have user scores—sourced from past and present behavior—allocate the majority of tokens to crowdsales that clear at different prices based on user scores. Better users get bigger allocations at lower prices. Mercenary farmers pay full price—or get no access at all. By requiring users to have skin in the game and giving them a cost basis, you create a more committed holder base rather than farmers looking to cash out free money. Crowdsales also add a built-in sybil resistance mechanism. Free money attracts noise. @clairekart is right that the airdrop meta emerged in response to regulation—in a free market, crowdsales are just a better way to distribute most tokens. Even Ethereum was distributed via crowdsale. With regulatory clarity finally emerging, why can’t your users be your "distributed BlackRock"? Your thousands of investors who've demonstrated they're long-term value-add holders. What should go into a "holder score"? It depends on the project, but some ideas: * Token retention curves (7/30/90/365-day holding percentages) * Governance participation * Fee spend * LP provision days * Relevant social engagement / Kaito scores * Product usage metrics, shit like that If you publish this in a standardized JSON format, others protocols can easily ingest and incorporate into their own distributions. It’s the same reason finance companies freely share data on their users to credit bureaus—users behave better with you when they know their reputation travels across platforms. So yes, airdrops are dumb, but they don’t have to be. Unless you're running pay-for-performance airdrops, if you have an airdrop at all, it should be small (<15% of total TGE). The remaining portion should be sold in score-tiered crowdsales, with pricing tiers published upfront so everyone knows the rules. (Be fully transparent, filter out team and investor addresses proactively.) And keep holder scores updated throughout subsequent campaigns and reward seasons. Now instead of rewarding people gaming the snapshot, you reward staying power and real users. IMO that will result in cleaner distributions, clearer PMF signals, and token holders who actually give a shit about your project, instead of dumpers who are hemorrhaging tokens over time. It's crypto—the design space is a lot bigger. Let’s use it. Disclosure: Dragonfly is an investor in several of the assets I mentioned, also I have done absolutely zero conferring with lawyers about this, so consider this a shower thought and definitely not legal advice!