Rollup Economics
Rollup economics describes the cost structure, revenue model, and profit dynamics of Layer 2 rollups that post data to a base chain.
Key Takeaways
- Rollup economics follows a straightforward formula: the sequencer collects fees from users, pays costs to post data and proofs to the base layer, and the difference is the rollup's margin.
- EIP-4844 (March 2024) transformed rollup profitability by introducing blob transactions with a separate fee market, cutting Layer 2 data posting costs by approximately 90% and making sub-cent user transactions viable.
- The cost structure differs between optimistic and ZK rollups: optimistic rollups spend primarily on data posting, while ZK rollups carry additional proof generation costs that can dominate their expenses.
What Is Rollup Economics?
Rollup economics refers to the financial model governing how Layer 2 rollups generate revenue, incur costs, and sustain operations. At its core, a rollup operates like any business: it provides a service (cheap, fast transaction execution) and charges users fees that must exceed its operating costs to remain viable.
Understanding rollup economics matters because it reveals the incentive structures that determine which rollups survive, how fees will evolve, and where value accrues in the blockchain stack. As Layer 2 networks handle an increasing share of transaction volume, their economic sustainability directly affects millions of users and billions of dollars in activity.
The economics shifted dramatically in 2024 when Ethereum's Dencun upgrade introduced blob transactions, slashing data costs and reshaping the profitability landscape for every major rollup.
How It Works
A rollup's economic model has two sides: revenue collected from users, and costs paid to the base layer and infrastructure providers. The sequencer sits at the center, acting as the entity that bundles user transactions, executes them off-chain, and posts results on-chain.
Revenue: User Fees
Rollup revenue comes from the fees users pay for transaction execution. These fees have two components:
- Execution fee: covers the computational cost of processing the transaction on the rollup's virtual machine. This is analogous to the gas fee on Ethereum but substantially cheaper since execution happens off-chain.
- Data fee (L1 fee): covers the rollup's cost of posting compressed transaction data to the base layer for data availability. This is typically the larger component and fluctuates with base layer congestion.
The sequencer estimates these costs and adds a margin when quoting fees to users. This margin, sometimes called the "sequencer surplus," is the rollup's gross profit.
Cost Structure
Rollup operating costs fall into four main categories, each with distinct dynamics:
| Cost Category | Description | Primary Driver |
|---|---|---|
| Data posting | Publishing compressed transaction data to the base layer via calldata or blobs | Base layer fee market conditions |
| Proof generation | Computing validity proofs (ZK rollups only) to verify state transitions | Batch size, circuit complexity, prover hardware |
| Proof verification | On-chain gas to verify submitted proofs | Proof system type (SNARK vs STARK) |
| Sequencer infrastructure | Servers, bandwidth, state storage for running the rollup node | Transaction volume, state size |
The Profit Formula
A simplified view of rollup profit per batch:
Rollup Profit = Total User Fees Collected
- L1 Data Posting Cost
- Proof Generation Cost (ZK only)
- Proof Verification Cost (ZK only)
- Sequencer Operating Cost
Where:
L1 Data Posting Cost = compressed_batch_size × L1_fee_per_byte
User Fee per Tx = (L1 cost / txs_in_batch) + execution_fee + marginThe critical insight: data posting costs are amortized across all transactions in a batch. A batch containing 1,000 transactions pays the same L1 fee as a batch containing 100 transactions, so higher throughput directly improves per-transaction economics. This creates a flywheel: lower fees attract more users, more users fill batches faster, fuller batches reduce per-transaction costs further.
The EIP-4844 Transformation
Before March 2024, rollups posted transaction data as calldata on Ethereum. Calldata competes in the same gas fee market as regular transactions, making data posting expensive during periods of congestion. Data costs often represented 80-90% of a rollup's total expenses, leaving thin margins.
EIP-4844 (proto-danksharding), activated with Ethereum's Dencun upgrade in March 2024, introduced blob transactions: a new transaction type carrying large data blobs (~128 KB each) with a completely separate fee market. Each Ethereum block can include up to 6 blobs, and the blob base fee adjusts independently of regular gas prices.
Before vs After EIP-4844
| Metric | Pre-EIP-4844 (Calldata) | Post-EIP-4844 (Blobs) |
|---|---|---|
| Average L2 transaction fee | $0.10 to $1.00+ | Under $0.01 |
| Data posting cost share | 80-90% of rollup expenses | Under 20% of rollup expenses |
| Rollup margin | Thin, volatile with L1 congestion | Wider and more predictable |
| Data medium | Calldata (permanent, expensive) | Blobs (pruned after ~18 days, cheap) |
The impact was immediate and dramatic. Within weeks of the Dencun upgrade, major rollups saw their L1 data posting costs collapse. Arbitrum and Optimism reported data costs dropping by over 95%. Base, Coinbase's rollup, saw per-transaction L1 fees fall from several cents to fractions of a cent. This translated directly to cheaper user fees and wider sequencer margins.
For a detailed analysis of the blob fee market dynamics, see the research article on EIP-4844 and the blob fee market.
The Blob Fee Market
Blob fees use the same EIP-1559 mechanism as regular gas: a base fee that adjusts up or down depending on utilization relative to a target. The target is 3 blobs per block (with a maximum of 6). When utilization exceeds the target, the blob base fee increases; when below target, it decreases. During periods of low demand, blob fees can fall to near zero, effectively making data posting almost free for rollups.
This separate fee market means rollup data costs no longer spike when NFT mints or DeFi events cause gas wars on Ethereum's execution layer. Rollups compete for blob space only with other rollups, creating a more predictable cost environment.
Optimistic vs ZK: Cost Comparison
The two main rollup architectures have fundamentally different cost profiles, which influences their economic viability at different scales.
Optimistic Rollup Costs
Optimistic rollups (Arbitrum, Optimism, Base) have a simpler cost structure because they skip proof generation entirely during normal operation. Their primary costs are:
- Data posting to Ethereum (via blobs): the dominant cost, now significantly reduced post-EIP-4844
- State commitment transactions: periodic on-chain submissions of the rollup's state root
- Fraud proof execution (rare): only incurred if someone challenges a batch, which almost never happens in practice
ZK Rollup Costs
ZK rollups (zkSync Era, Starknet, Linea, Scroll) carry the same data posting costs plus significant proving overhead:
- Data posting to Ethereum (via blobs): same as optimistic rollups
- Proof generation: running the prover (specialized hardware or cloud GPU clusters) to generate validity proofs. This can cost $0.01 to $0.10 or more per proof depending on the circuit complexity and hardware
- Proof verification on-chain: submitting and verifying the proof on Ethereum costs gas. SNARK verification typically costs 200,000-300,000 gas, while STARK verification can cost more due to larger proof sizes
The trade-off: ZK rollups spend more on proving but gain faster finality (no 7-day challenge period) and can potentially compress data more aggressively since validity proofs guarantee correctness without requiring full transaction data on-chain.
Sequencer Revenue and MEV
The sequencer occupies a privileged position in rollup economics. As the sole entity ordering transactions, it has the power to extract additional value beyond the stated fee margin.
Sequencer Extractable Value
Sequencer extractable value (SEV) is the rollup equivalent of MEV on Layer 1. Because centralized sequencers control transaction ordering, they can:
- Front-run user transactions by inserting their own trades before profitable swaps
- Back-run transactions by placing trades immediately after large swaps to capture arbitrage
- Sandwich attack DEX trades by placing orders on both sides of a user transaction
- Prioritize transactions from users willing to pay higher tips
Whether and how much MEV a sequencer extracts varies by rollup. Some rollups have adopted fair ordering mechanisms or committed to not extracting MEV as a competitive differentiator. Others route MEV revenue to the rollup's treasury or DAO. The lack of transparency around sequencer behavior makes it difficult to quantify SEV across the ecosystem.
Priority Fee Revenue
Similar to Ethereum's priority fees, rollups may allow users to pay extra for faster inclusion. During periods of high demand, this creates an additional revenue stream for the sequencer beyond the base fee margin.
Why It Matters
Rollup economics determines who pays, who profits, and whether Layer 2 solutions are sustainable long-term. For users, understanding these dynamics explains why fees vary across rollups and over time. For builders, it reveals the competitive landscape: rollups with better economics can offer lower fees, attract more users, and build stronger network effects.
The economics also highlight a tension in blockchain scaling. When rollup data costs drop (as they did with EIP-4844), users pay less but the base layer earns less fee revenue. This has implications for Ethereum's security budget and the long-term sustainability question of whether base layers can remain secure if most economic activity migrates to Layer 2s.
For Bitcoin-based scaling approaches, rollup economics offers a useful comparison point. Solutions like Spark use different architectures (statechains rather than rollups) that avoid the data posting costs entirely, since they do not require posting compressed transaction data to block space on the base layer for every batch.
Use Cases
- Fee benchmarking: comparing the true cost of transacting across different rollups requires understanding each one's margin and data cost structure
- Investment analysis: rollup token valuation models depend on understanding sequencer revenue, operating costs, and how fee structures may evolve with scaling upgrades
- Protocol design: new rollups use economic modeling to determine optimal fee structures, batch sizes, and proving schedules that balance user cost with sustainability
- Scaling roadmap evaluation: understanding rollup economics clarifies which scaling approaches (rollups, state channels, sidechains) are viable for different use cases and transaction volumes
Risks and Considerations
Sequencer Centralization and Rent Extraction
Most rollups today run a single centralized sequencer. This means one entity controls fee pricing, transaction ordering, and margin extraction. While competition between rollups provides some price discipline, users on a specific rollup have limited recourse if the sequencer raises fees or begins extracting MEV more aggressively. Decentralized sequencer designs are under active development but remain unproven in production.
Fee Volatility from L1 Congestion
Despite EIP-4844's improvements, rollup fees still correlate with base layer conditions. If blob demand exceeds supply (all rollups competing for 6 blobs per block), blob fees will rise sharply. As more rollups launch and blob demand grows, this could recreate the cost pressure that existed before EIP-4844, until full danksharding expands blob capacity further.
Race to Zero Fees
Competition among rollups puts downward pressure on fees, which benefits users but threatens rollup sustainability. If margins compress too far, rollups may struggle to fund development, security audits, and infrastructure. Some rollups may resort to MEV extraction or token subsidies to maintain operations, creating hidden costs for users.
Value Accrual Uncertainty
It remains unclear where value ultimately accrues in the modular blockchain stack. If rollups commoditize and compete primarily on fees, most value may flow to the base layer (for security) or to applications built on top (for user attention). Rollup token holders face the risk that sequencer revenue alone may not justify current valuations.
This glossary entry is for informational purposes only and does not constitute financial or investment advice. Always do your own research before using any protocol or technology.