Since 2020, the crypto industry has become increasingly focused on “exchanges and entrepreneurship,” forming a closed利益 loop among project teams, VCs, exchanges, and “farming studios.” False data drives out genuine users, airdrop mechanisms have turned into blood transfusion systems feeding bots, and the industry is trapped in a dilemma where bad money drives out good. This article is based on a piece by danny,整理、翻譯及撰稿由 PAnews 完成。

(Previous context: Data: Nearly 85% of 118 tokens issued this year at TGE failed to launch successfully. Can you still profit from new listings?)

(Additional background: Why are ICOs regaining dominance in on-chain fundraising? The three underlying reasons that beat Airdrops.)

Table of Contents

• The incentive structure of “false” economy: from value creation to alienation solely for listing
  • 1.1 Gatekeeper effect of exchanges: data as entry ticket
  • 1.2 VC pressure cooker: vanity metrics and exit liquidity
  • 1.3 Alienation of marketing activities: from customer acquisition to feeding bots
• Industrialized operation mechanisms of “farming studios” (supply side analysis)
  • 2.1 Industrial-grade infrastructure and automation
  • 2.2 “Task platforms”: training grounds and conspirators for industrialized volume farming
  • 2.3 Economics ledger of farming: ROI-driven capital allocation
• Ruins beneath data surface: token issuance. People gone. Buildings empty.
  • 3.1 Starknet: avalanche of retention rates and extremely high customer acquisition costs
  • 3.2 zkSync Era: “Era” ending and data cliff
  • 3.3 LayerZero: self-reporting mechanism triggering community wars and trust crises
• The phenomenon of bad money driving out good in digital assets
  • 4.1 Several methods of exclusion mechanisms
  • 4.2 “Noise floor” and signal loss
  • 4.3 Project teams’ “intoxication” and collusion
• Conclusion

Around winter 2020, project teams’ goals shifted from “creating value and serving users” to “listing on exchanges and pleasing studios.” The core driver behind this phenomenon is the contradiction between exchanges’ rigid data requirements and the cold start of early projects. Due to the lack of real initial users and data, but with exchanges needing this data, project teams are forced to collude with studios, using volume farming to create false prosperity and meet market expectations.

This model leads to a dual focus: “To Exchange” (面向交易所創業) and “To Airdrop Hunter” (面向空投獵人創業). Under this background, the industry exhibits a “bad money drives out good” phenomenon, where false, arbitrage-focused interactions (bad money) crowd out genuine, utility-oriented users (good money). By diluting rewards and raising costs, real users are driven out.

Initially, airdrops were designed as marketing tools to attract new users and trigger network effects. But their original purpose has been completely undermined, turning into blood transfusions for studios and bots. Both project teams and exchanges are immersed in this data façade, leading to massive resource waste and fundamentally misleading the industry’s development direction.

This article aims to explore the root causes, mechanisms, and future impacts of this phenomenon. We will examine how top-tier exchanges like Binance and OKX, through their listing standards, inadvertently become “conductors” of this distorted incentive mechanism; analyze how VCs, with “high FDV and low circulating supply” tokenomics, form a covert symbiosis with farming studios, jointly orchestrating this false prosperity.

“False” economy’s incentive structure: from value creation to alienation solely for listing

The proliferation of farming studios is not accidental chaos but a rational economic response to the current crypto market’s established incentive structure. To understand why project teams even “tacitly accept” the existence of studios, we must first analyze the survival rules set by the industry gatekeepers—CEXs, VCs, and KOLs.

1.1 Gatekeeper effect of exchanges: data as entry ticket

In the current token economy model, for most infrastructure and middleware protocols, a successful “full house” listing on top exchanges (like Binance, OKX, Coinbase) is the definition of project success. This is not only a necessary liquidity event for early investors to exit but also a mark of mainstream market recognition. However, the listing standards of exchanges objectively create a demand for false data.

Exchanges rely on quantitative indicators for review. Binance, as the largest market share exchange, emphasizes “strong community support” and “sustainable business model,” but in practice, metrics like trading volume, daily active addresses, on-chain transaction count, and TVL are heavily weighted. OKX also explicitly states that besides technical factors, they focus on “adoption rate” and “market position.”

This mechanism creates a classic “cold start paradox”: a new Layer 2 or DeFi protocol needs real users to qualify for listing, but without the liquidity and token incentives from listing, attracting genuine users is difficult. Farming studios fill this vacuum—they provide a “growth-as-a-service” solution. Using automation scripts, studios can generate tens of thousands of daily active addresses and millions of transactions in a short time, drawing a perfect growth curve to meet the data requirements of due diligence teams.

This pressure is also reflected in the rumored “listing fee.” Although top exchanges like Binance deny charging high fees and emphasize transparency, in reality, project teams often need to promise certain trading volume liquidity or provide large amounts of tokens for marketing budgets. If the project lacks natural traffic, they must rely on market makers and studios to maintain this false prosperity, avoiding delisting or being placed under observation.

1.2 VC’s pressure cooker: vanity metrics and exit liquidity

VCs play a catalytic role in this ecosystem. In past cycles, billions of dollars flowed into infrastructure tracks. Their business model requires seeking exit paths. The standard lifecycle of a crypto project includes seed round, private placement, TGE, and listing.

During TGE, valuation and market buzz are highly correlated. Due to the lack of traditional P/E or cash flow models, valuation relies on proxy indicators:

• Active addresses are interpreted as “users.”
• Transaction count as “demand for on-chain space” and “user activity.”
• TVL as “trusted capital” and “cold start funding.”

Influenced by industry cleansing and the myth of quick wealth, crypto attracts many short-term speculators who prioritize these “soil indicators” over substantive value. VCs know they are competing with retail for limited liquidity, so they push their portfolio companies to maximize these metrics before TGE.

This creates a serious moral hazard: VCs are incentivized to overlook or even secretly promote Sybil activities, because these studios’ data support their high valuation exits. Hence, some TGE projects’ Twitter accounts boast nearly a million followers, with address interactions approaching hundreds of millions, and transaction counts reaching billions.

While total registered users or raw transaction volume seem convincing on the surface, they often lack relevance to long-term business success. Yet, at the negotiation table of primary markets, these metrics are standard entry conditions. A project with 500,000 “active addresses” (even if 99% are bots) is valued far higher than one with 500 genuine high-net-worth users.

1.3 Alienation of marketing activities: from customer acquisition to feeding bots

Originally, airdrops aimed to be decentralized marketing tools, distributing tokens to real users to trigger network effects. But under current incentive structures, the nature of airdrops has fundamentally changed.

Project teams find it more cost-effective to attract studios through implied airdrop expectations rather than spend budgets on educating the market and finding genuine users (a slow and expensive process). This “points system” or “mission-based” marketing is essentially a data purchase transaction (some call it a forward discount token purchase). The project pays (or promises to pay) tokens, and studios deliver on-chain data, gas fees, and transaction costs. This short-term mutually beneficial transaction allows the project to showcase attractive data to exchanges and VCs, while studios gain expected token rewards.

However, this collusion harms the entire industry’s product culture and genuine users. Studios only need to meet minimal interaction thresholds (e.g., once a week, over $10), and the project’s product iteration begins to revolve around these bots and scripted interactions rather than improving real user experience. This leads to the birth of many “zombie protocols” with no real utility—designed solely for bots. Come on, who would cross-chain to swap a $10 token just for fun?

Farming studios’ industrialized operation mechanisms (supply side analysis)

The term “farming studio” carries a grassroots connotation, even some internet jokes, as a form of self-deprecating humor from the community. But in the 2024-2025 context, it refers to a highly professionalized, capitalized, and even software-capable high-tech industry. These entities operate with the efficiency of software companies, utilizing complex tools, sophisticated algorithms, and infrastructure to maximize reward mechanism exploration.

2.1 Industrial-grade infrastructure and automation

The barrier to participating in Sybil attacks has been significantly lowered, thanks to the proliferation of professional tools. Fingerprint browsers like AdsPower and Multilogin allow operators to manage thousands of independent browser environments on a single machine. Each environment has a unique digital fingerprint (User Agent, Canvas Hash, WebGL data, etc.) and a separate proxy IP. This makes traditional anti-cheat measures based on device detection (e.g., login from the same device) completely ineffective.

A typical studio operation includes several highly industrialized steps:

Identity disguise and isolation: Using fingerprint browsers to isolate local storage and cookies for thousands of wallets, ensuring they appear as unrelated users from around the world.

Batch wallet generation and management: Using hierarchical deterministic (HD) wallets to generate addresses in bulk. To evade on-chain clustering analysis, studios often use CEXs supporting sub-accounts for fund distribution. Since CEX hot wallet addresses are shared, this cuts off the link to on-chain fund sources, breaking common “witch-hunter” tracking graphs. (Advanced versions also stagger transfer times, amounts, etc.)

Scripted interaction execution: Writing Python or JavaScript scripts, combined with Selenium or Puppeteer automation frameworks, to run on-chain interactions 24/7. These scripts can automatically perform swaps, bridges, lending, and even incorporate randomness to simulate human-like intervals and amount fluctuations, fooling behavior analysis algorithms.

KYC supply chain: For projects attempting to block studios via enforced KYC (e.g., CoinList public offerings or certain project verifications), an underground KYC data industry has formed. Studios can cheaply buy real identity info and biometric data from developing countries in bulk, even using AI to pass liveness detection, thoroughly breaking the Proof of Personhood defense.

2.2 “Task platforms”: training grounds and conspirators for industrial volume farming

Another key development this cycle is that besides Web3 task platforms like Galxe, Layer3, Zealy, Kaito, and the like, legitimate wallets and project teams—such as Binance alpha, various Perp DEXs, and emerging L1s—have joined the ranks. These platforms claim to be tools for user education or community building, issuing “tasks” (e.g., “Cross-chain ETH to Base,” “Trade once on Uniswap”) to reward users with points or NFTs.

In reality, these platforms serve as “training grounds” and “task lists” for farming studios.

Layer3 operates a “growth-as-a-service” marketplace. Protocols pay Layer3 for traffic, and Layer3 distributes these tasks to users. For studios, Layer3 clearly lists interaction paths approved by project teams. They only need to script these specific paths to obtain “officially certified” interaction records at minimal cost.

Kaito is another media buy service market, filled with AI bot voices, indirectly fueling the proliferation of AI-generated comments and spam on Twitter.

Galxe allows projects to create tasks involving on-chain interactions and social media follows. While Galxe offers some anti-witch features (like Galxe Passport), these are often paid premium options, and many projects deliberately keep strict filtering off to maximize participation data.

Ironically, these platforms inadvertently (or perhaps intentionally) train bots. By standardizing complex interactions into linear “Task A + Task B = Reward,” they create a deterministic logic that scripts excel at handling. The result is a large number of “mercenary users” who mechanically perform the minimum actions needed for rewards, stopping immediately after completing tasks.

2.3 Economics ledger of farming: ROI-driven capital allocation

The essence of farming studios is capital allocation strategy. In their ledgers, gas fees, slippage losses, and capital occupation costs are viewed as customer acquisition costs. They calculate the return on investment (ROI).

For example, spending $100 in gas fees on a cluster of 50 wallets, and ultimately obtaining $5,000 worth of airdrop tokens, yields an ROI of 4,900%. Such huge profits are not rare in history:

Starknet case: a typical GitHub developer account can receive about 1,800 STRK tokens. During the initial token release, the price exceeded $2, meaning a single account could earn over $3,600. If a studio scripts and maintains 100 GitHub accounts, total gains could surpass $360,000.

Arbitrum case: the Arbitrum airdrop distributed about 12.75% of total tokens. Even wallets with minimal interaction records received tokens worth thousands of dollars. This massive liquidity injection not only proved the feasibility of the studio model but also provided ample capital for larger-scale attacks in future cycles (zkSync, LayerZero, Linea).

This high ROI creates a positive feedback loop: successful airdrops fund studios, enabling them to develop more sophisticated scripts, buy more expensive fingerprint tools and proxies, and further dominate future projects, squeezing out genuine users.

The ruins beneath data surface: token issuance. People gone. Buildings empty.

The “victory” of studios is starkly reflected in the poor performance of major protocols after airdrops. This reveals a clear pattern: manufactured growth -> snapshot -> retention collapse.

3.1 Starknet: avalanche of retention and extremely high customer acquisition costs

Starknet, a highly anticipated ZK-Rollup network, launched a large-scale STRK token airdrop in early 2024. Its distribution was broad, targeting developers, early users, and ETH stakers.

The data is shocking. On-chain analysis shows that only about 1.1% of addresses that claimed the airdrop remained active afterward. This means 98.9% of the profit addresses are mercenaries, who stop contributing immediately after collecting rewards.

Starknet spent roughly $100 million (based on token value) to acquire about 500,000 users. But with a 1.1% retention rate, the cost per retained user skyrockets above $1,341. For any Web3 protocol or Web2 company, this is an economically unsustainable disaster.

This pressure caused STRK token price to plummet 64% after launch. Although the total market cap seemed to grow due to token unlocks, the tokens’ purchasing power was greatly diminished.

Starknet’s case offers a textbook example: users “bought” through airdrop expectations are just illusions. Studios extract value and shift to the next battlefield, leaving protocols with inflated historical data and empty on-chain space.

3.2 zkSync Era: “Era” ending and data cliff

zkSync Era’s trajectory mirrors Starknet. Before the snapshot, activity on the network surged exponentially, often surpassing Ethereum mainnet, hailed as the L2 leader.

But after the airdrop announcement and snapshot date confirmation, activity on zkSync Era collapsed immediately. The 7-day average active addresses dropped from a peak of 455,000 in late February 2024 to 218,000 in June—a 52% decline. Daily transaction count plummeted from 1.75 million to 512,000. Notably, this decline occurred before token distribution.

Nansen data shows that nearly 40% of the top 10,000 wallets received all their tokens and sold them within 24 hours. Only about 25% chose to hold.

This pre-distribution activity drop confirms that the previous boom was entirely driven by external incentives. Once the “snapshot” was perceived as completed by studios, they immediately stopped scripts. The data decline is superficial; the real truth is the collapse of the “ecological prosperity” narrative.

3.3 LayerZero: self-reporting mechanism triggers community wars and trust crises

Cross-chain interoperability protocol LayerZero attempted a radical move against studios: launching a “self-report” mechanism. Project teams proposed a transaction: if you admit to being a witch, you can keep 15% of the airdrop; if you hide and are caught, you get nothing.

LayerZero ultimately identified and marked over 800,000 addresses as potential witch attackers. This strategy caused a huge rift in the community. Critics argued that labeling users who used “minting tools” like Merkly as witches was unfair, as LayerZero had previously benefited from these users’ cross-chain fees and transaction data.

Although this “purge” redistributed tokens to “loyal users,” LayerZero’s price still fell 23% within a week after listing. More seriously, the “witch bounty hunter” plan led to mutual reporting among community members, creating a toxic surveillance and confrontation atmosphere, severely damaging the project’s reputation.

$ZRO The phenomenon of bad money driving out good in digital assets

In economics, when exchange rates are fixed, bad money drives out good. In the context of crypto user acquisition, this manifests as: false users displacing real users.

4.1 Several exclusion mechanisms

Reward dilution: Airdrops are often zero-sum games. Projects allocate a fixed percentage (e.g., 10%) of tokens to the community. If a studio controls 10,000 wallets, they cut a huge slice from the pool, greatly diluting the share of genuine users with only one wallet. When real users realize that a year’s normal usage yields negligible rewards, their willingness to participate plummets.

Network congestion and fee spikes: Industrial volume farming consumes precious block space. During peak farming periods (e.g., Linea Voyage or Arbitrum Odyssey), gas fees soar. Genuine users, unable to afford high transaction costs, are forced to migrate to other chains or stop using. Eventually, only bots remain—since bots can amortize high airdrop returns against high gas costs, while real users’ utility cannot cover these costs.

Complex mechanisms: Some TGE projects intentionally design interaction tasks to be extremely complex to block bots, but the complexity itself deters natural users, leaving only tireless bots capable of completing them. Interestingly, some commentaries suggest that by 2025, the Perp DEX wars have evolved into script wars.

4.2 “Noise floor” and signal loss

The proliferation of studios raises the entire ecosystem’s noise floor. With 80%-90% of traffic being inorganic, project teams cannot discern true product-market fit.

In such high-intensity data pollution and toxic transaction environments, traditional A/B testing, user feedback loops, and feature adoption metrics become useless. Ultimately, projects optimize UI/UX based on script preferences (e.g., reducing clicks for easier script operation, not for human usability).

The market falls into a “Market for Lemons” dilemma. High-quality projects that refuse to farm are undervalued; while low-quality projects that actively farm and generate “hot” data attract capital and attention. This leads to high-quality projects being forced out or colluding, degrading overall market quality.

4.3 Project teams’ “intoxication” and collusion

Under the influence of the macro environment and exchange tacit approval, some project teams begin to “drift” in data façade. Beautiful data is the only credential they can show to investors and the public. Admitting that 90% of users are fake would cause valuation collapse, possibly preventing listing and risking lawsuits.

Thus, project teams fall into a “performative ignorance” state. They implement seemingly strict anti-bot measures (e.g., banning low-level scripts) but deliberately leave “backdoors” for high-level studios. Layer3’s co-founder even publicly admitted that some projects prefer not to conduct strict bot filtering because they are optimizing metrics that drive narratives and fundraising.

This collusion completes a closed loop—project teams need fake data to sell to VCs/exchanges; studios provide fake data to project teams; VCs/exchanges package projects and sell to retail.

Conclusion

The current industry resembles an athlete on too many stimulants (false data): short-term muscle growth (TVL, user count) but internal organs (real income, community consensus) are exhausted.

It was once a path to change the world as cyberpunk, but the crypto ecosystem has degenerated into a Performative Economy, where projects pay fees or sign options with studios to “produce” data that meets the arbitrary standards of exchanges and VCs.

It’s not that studios are doing wrong—after all, it’s business. Demand creates supply. But when the entire market is flooded with studios and incentive traces of traffic, things change.

This "project-VC-exchange-studio"利益閉環 is a typical zero-sum game. It sustains short-term prosperity by consuming industry trust reserves. To break this vicious cycle, the industry must undergo a painful “deleveraging” process.

For project teams, chasing exchange listing standards has replaced exploring product-market fit (PMF). Projects are designed to be “farmed” rather than “used.” Moreover, hundreds of billions of dollars in token incentives—originally meant to activate genuine communities—are siphoned off by professional extraction and arbitrage, ultimately abandoned.

This is not just bad money driving out good, but false money displacing real. Unless the industry shifts focus from vanity metrics like “active addresses” and “transaction counts” to attracting real usage and creating genuine economic value, we will only go further down the path of bad money driving out good.

Farming studios won the airdrop battle, but their victory may cause the crypto industry to lose the war for mass adoption.

Perhaps only when “product usage” yields more than “data farming” can good money return, and the crypto industry truly emerge from the quagmire of false prosperity into the realm of real technology implementation.

By 2026, may we be humble players in this “data-driven” era.