Airdrops — the retro system, 10-20 accounts, a process

Introduction

An airdrop is a distribution of tokens for early use of a protocol. Between 2020 and 2024 it was the largest source of asymmetric return in crypto: Uniswap, ENS, Arbitrum, Optimism and Celestia handed out 5-6 figure sums for the cost of gas. In 2024-2025 most projects switched to a points model, anti-Sybil checks and long vesting schedules. The airdrop did not die — it became operational work.

This lesson teaches you how to build a system for farming retroactively across 10-20 accounts, with cost, time and risk measured, without being disqualified as a Sybil. This is not a lottery. It is a process, like running small product tests with quality control.

The logic of an airdrop — why projects give tokens away

A project hands out tokens because:

1. It decentralises supply (regulations, DAO)
2. It buys users more cheaply than advertising
3. It builds TVL and metrics ahead of the next VC round

The criteria are usually secret until the end, but they are made up of:

Projects use anti-Sybil filters: Nansen, Gitcoin Passport, funding-graph analysis (whether 20 accounts were funded from a single address).

The 10-20 account architecture — where to start

Do not start with 100 accounts. Start with 5, refine the process, then scale to 10-20.

Structure

• Every wallet gets its own seed (BIP39), not 20 addresses from one seed — those are easy to link.
• Funding: source -> CEX -> split into different amounts and time intervals -> bridging through different routes.
• Activity: not identical. Introduce human noise.

Funding (anti-Sybil scheme):
CEX sub-account A -> 0.12 ETH -> Stargate bridge -> Wallet 1 (Base)
CEX sub-account A -> 0.11 ETH 2 days later -> Across bridge -> Wallet 2 (Base)
CEX sub-account B -> 0.13 ETH -> Wallet 3 (Arbitrum) -> official bridge -> Base

Never:
One address -> send 0.1 ETH to 20 wallets within 10 min along the same route -> Sybil flag.

Cost of entry and ROI

Cost of farming project X:
Cost = Gas (entry + interactions + exit) + bridge spread + Time * hourly rate

ROI = (Expected airdrop per account * Number of accounts * Probability of qualifying) - total cost

Example:
Gas $15 per account * 20 = $300
Time 5h/month * 3 months = 15h
Expected airdrop $400 per account * 20 * 50% qualification = $4000
ROI = $4000 - $300 - cost of time

You estimate the expected airdrop from earlier airdrops in the same narrative (L2s: $300-800, bridges: $100-400). Do not count on more than history has shown.

The operating system — calendar and SOP

You need an SOP (Standard Operating Procedure) for every project.

An SOP consists of:

1. Entry criteria: Did the project raise >$10M, does it have a public team, a roadmap longer than 6 months? You do not farm every testnet.
2. Weekly task list: e.g. for an L2: 1 swap, 1 LP, 1 bridge, 1 niche contract per week.
3. Calendar: Mark snapshots, Galxe quest deadlines, points.
4. Tracking: A spreadsheet with the account list, the date of the last activity and the gas spent.

Tools: DeBank portfolio tracker for many accounts (watch mode), Arkham for tracking gas, your own Google Sheet.

Diversifying activity (humanising it)

Bots do the same thing every hour. A human does not.

• Different amounts: one swap of $10, another of $27, a third of $54.
• Different intervals: one account runs its tasks on Monday, another on Thursday.
• Different dApps: not every account swaps the same token on the same DEX.
• Leave “loyalty” fingerprints: a DAO vote, a transaction on a bad day, a second interaction two weeks later, not only on quest day.

The 2024-2025 points narrative and how to count points

Most new projects (EigenLayer, deBridge, Hyperliquid, etc.) use points as a proxy for the airdrop. Points = a promise, not a guaranteed conversion rate.

Rules:

1. Do not look at the points leaderboard. Look at points per $ of capital and points per unit of time.
2. Check whether points are linear or decaying — some projects reward earliness.
3. Can points be lost? (e.g. withdrawing from the pool).
4. What is the FDV target and what percentage goes to the community? From that, calculate the implied value of a point.

Implied value of a point:
Estimated FDV * airdrop allocation % * (your points / total points) = estimated value
Example: FDV $1B, allocation 10% = $100M, your points 10k / 10M total = 0.1%
Value = $100k -> but total points grow and FDV is uncertain, so discount 70%
Realistically: $100k * 0.3 = $30k * probability of TGE

Anti-Sybil — when you get cut out

Projects are getting better at detecting farms. The following gets flagged:

• The same funding source
• The same amounts, the same contracts in the same order
• Activity only on quest days
• No mainnet activity, only a farmed L2
• 20 wallets voting from the same IP on Gitcoin Passport (do not create 20 passports from one IP)

Mitigation: split your funding sources, vary amounts and timing, keep a history on 2-3 base chains, and use genuine interactions (if you use a bridge, actually use it, not only to farm points).

Remember: breaking a project’s terms, which state “1 person = 1 account”, can lead to disqualification and lost time. Judge whether the project allows multiple wallets per entity.

When the plan says do not trade (do not farm)

1. When a project has no funding, no team and no audit, yet asks you to sign an unlimited approval. The risk of a drain is greater than the potential airdrop.
2. When the cost of gas + bridge > 20% of the expected reward per account. Negative EV across 20 accounts is a certain loss. Farm the L2.
3. When you have to buy the project’s token to farm its airdrop, and the token unlocks in 7 days. You are buying supply exposure in order to get an airdrop — double risk.
4. When the system requires KYC on every account and linking your ID. That is not farming, it is selling your identity. The plan says no.
5. When a project changes its scoring rules three times along the way. No predictability = little respect for the farmer’s time. Move your time elsewhere.

The most expensive mistakes

1. Scaling to 50 accounts from the start without an SOP. You lose seeds, mix up accounts, send from the wrong wallet, burn gas and time. Start with 5.
2. Funding every account from one address in a single day. A perfect Sybil graph for Nansen. You hand the project a free reason to cut you out.
3. No cost spreadsheet. After 6 months you have no idea how much you spent on gas. You think you are “definitely up”. Measurement > gut feeling.
4. Using the same IP and browser fingerprint for every account with no separation. If the project uses browser-fingerprint analysis, one ID links all your accounts.
5. Holding the airdrops you earned without a plan. An airdrop is a reward that, in 70% of cases, drops 60% within 30 days of the TGE. If you have no sell plan (e.g. DCA out 30% immediately and 70% over 2 weeks), you are back to square one.

A measurable exercise

Building the system in 7 days:

Day 1-2: Pick 2 projects (e.g. one L2, one bridge). Check the funding on CryptoRank, the TVL on DeFiLlama, and the audit.

Day 3: Create 5 burner wallets following the procedure. Write the seeds down on paper, build the tracking spreadsheet.

Day 4: Fund the accounts from a CEX with different amounts and routes, leaving a 24h gap between batches.

Day 5-7: Run the SOP: at least 3 different interactions per project, different amounts.

Measurement after 30 days:

Total cost = gas + bridge fee
Time = minutes spent * your rate
Number of unique contracts touched per account
Did any wallet get a Sybil flag (check Gitcoin Passport, the LayerZero checker)?

After 30 days, assess: how many projects met the criteria to continue? How many to cut?

Journal format:

| Date | Project | Wallet | Action | Amount | Gas | Tx link | Notes (e.g. bug) |

Test yourself

1. Describe 3 criteria a project may use to qualify for an airdrop.
2. Why is funding 20 accounts from one address with the same amount risky?
3. How do you calculate the ROI of farming for 10 accounts?
4. What is humanising activity and what is it for?
5. How do you estimate the implied value of a point?

Frequently asked questions

How many accounts is optimal? 10-20 for an individual with 5-8h a week. Above 20 you need automation, and automation without an SOP increases the risk of errors and Sybil flags. Start with 5 and measure how long one task cycle takes you.

Do I have to use different IPs for the accounts? You do not need 20 VPNs, but do not run every task from one IP in the same minute. Spread them out over time. If the project has a strict anti-Sybil policy (e.g. Starknet), use separate browser profiles and consider a mobile hotspot for some accounts.

Do testnets make sense? Yes, but only if they have a confirmed airdrop or are part of a campaign with points. A testnet with no gas cost is cheap, but time costs too. Choose testnets of projects with >$20M in funding and an active team on GitHub.

What about taxes on airdrops? In Poland an airdrop is usually treated as income at the moment of receipt (the market value at the time of receipt). Selling it later is a separate event. Keep records: date, value in PLN, quantity. Consult an adviser — this is not tax advice.

Farm on mainnet or on an L2? An L2 has cheaper gas, so better EV for small accounts. Mainnet carries history and prestige, but a $50 cost per cycle kills small accounts. Optimally: 70% of activity on L2s, 30% on mainnet (so you do not look like a pure Sybil farmer only on cheap chains).

---