Why Paymaster Subsidies Are Economically Unsustainable

Subsidizing gas is a classic CAC > LTV trap. Protocols spend real capital to acquire users who generate zero protocol revenue, creating a negative-sum game. This model only works with infinite VC funding.

• Unit Economics: Subsidies scale linearly with usage, not value captured.
• Exit Strategy: The only 'exit' is rugging users by removing subsidies, which destroys the product's core value proposition.

True account abstraction isn't about hiding gas, it's about flexible sponsorship. Protocols like Ethereum's EIP-4337 and zkSync's native AA enable sponsorship logic, not just blanket subsidies. The current subsidy model confuses user acquisition with infrastructure.

• Real AA: Allows for conditional, application-specific fee logic (e.g., 'sponsor only first tx').
• Fake AA: A simple relay that eats venture capital to mask transaction costs.

The viable model is session keys with explicit sponsorship windows, as seen in gaming or social apps. A dApp pays for a user's gas for a specific, high-value action (e.g., a game session, a trade), baking the cost into its own business model.

• Clear ROI: Sponsorship is a direct customer acquisition cost for a known conversion event.
• No Leakage: Sessions are time-boxed and scope-limited, preventing infinite subsidy drains.

The endgame is verifiable business logic for sponsorship. Using ZK proofs, a dApp can prove to a paymaster that a user's action will generate sufficient protocol fee revenue to cover the gas cost, enabling algorithmic, profit-positive subsidies.

• ZK Business Logic: Prove a trade will generate >X fees before sponsoring its gas.
• Automated Sustainability: Turns subsidies from a cost center into a growth engine with positive unit economics.

Protocols like Polygon and Biconomy initially subsidized gas to attract users. This creates a race to the bottom where the only differentiator is who burns more VC money. Once subsidies end, user activity evaporates, revealing no sustainable demand.

• Zero-Sum Game: Subsidies don't create new economic activity, they just shift it.
• Uncapturable Value: The subsidy cost is a direct loss, not a marketing expense with future ROI.

In a transparent mempool, subsidized transactions are a free signal for extractive value. Bots front-run, sandwich, and arbitrage the subsidized flow, capturing the value meant for the end-user. The protocol pays, the searcher profits.

• Adversarial Environment: Web2's benign user assumption is invalid on-chain.
• Economic Leakage: The intended subsidy is siphoned off by the MEV supply chain before reaching its target.

Subsidies require massive, idle capital pools (e.g., Starknet's 50M STRK allocation). This capital is locked, unproductive, and exposed to token volatility. It's a balance sheet liability, not an asset, creating massive opportunity cost and treasury risk.

• Sunk Cost Fallacy: Capital is spent, not deployed for yield.
• Tokenomics Contamination: Subsidy programs often inflate token supply or drain treasuries, harming long-term token health.

The sustainable model is not paying for gas, but facilitating transactions that create measurable economic surplus. Protocols like UniswapX and Across use intents and auctions to have users or fillers bear costs, only intervening to capture a slice of newly created value.

• Value-Based Pricing: Fees are a percentage of the value enabled, not a cost to be covered.
• Demand-Side Economics: The user or their counterparty pays because the transaction is net-positive for them.

Subsidies are a customer acquisition cost masquerading as a product. Protocols like Base and zkSync Era have spent hundreds of millions to bootstrap users, but this creates a perverse incentive where the most active users are the most expensive to retain.\n- Unit economics are negative: Cost per transaction > revenue per transaction.\n- Leads to mercenary capital: Users chase the next subsidized chain, offering zero loyalty.\n- Creates a 'subsidy cliff': Removing subsidies triggers a massive drop in activity, as seen in early L2 cycles.

Shift from blanket subsidies to programmatic sponsorship for specific, high-value actions. This aligns with the intent-centric architecture of UniswapX and CowSwap. The paymaster becomes a strategic tool, not a cost center.\n- Sponsor only profitable actions: A DEX pays gas for swaps that generate protocol fee revenue.\n- Enable new business models: Pay for user onboarding to your app's specific funnel.\n- Integrate with AA & session keys: Bundle sponsored transactions into a seamless user session, improving UX without blank checks.

Sustainable models require the protocol's own economic activity to fund its operations. This mirrors the evolution from VC-funded incentives to protocol-owned liquidity (POL) in DeFi.\n- Fee switch mechanics: A portion of protocol revenue funds a gas subsidy treasury, creating a flywheel.\n- Stablecoin gas tanks: Projects like Ethereum's EIP-4337 enable users to pay with any token, but the protocol can maintain a war chest of ETH/stablecoins to cover gas.\n- The real metric is LTV/CAC: The lifetime value of an acquired user must exceed the cost of their aggregate gas subsidies.