Layer 2 Has Scaled. Can It Unite?

Ethereum’s scaling problem has changed. A few years ago, the obvious constraint was price: routine transactions could cost tens or even hundreds of dollars when demand for blockspace surged. By 2026, many Layer 2 transactions cost fractions of a cent, while even Ethereum’s main network has become considerably cheaper during normal periods.

The technical achievement is substantial. Layer 2 networks process transactions away from Ethereum’s primary execution layer, compress the results and use the main blockchain for settlement or data verification. This allows more activity to be handled without requiring every Ethereum validator to reproduce every individual transaction.

Yet lower fees have exposed a second, more complicated problem. Ethereum is no longer one congested network. It is becoming an archipelago of rollups, application-specific chains and commercially operated ecosystems, each with its own liquidity, governance, bridge arrangements and security assumptions.

For businesses, investors and financial institutions, the central question is therefore no longer whether Layer 2 technology can make blockchain transactions cheaper. It is whether the emerging collection of networks can offer the predictability, interoperability and accountability required for serious financial activity.

Rollups have altered Ethereum’s economics

A Layer 2 rollup executes transactions outside Ethereum’s main chain and submits enough information to Ethereum for the results to be checked or reconstructed. Hundreds or thousands of user transactions can be grouped into one compressed submission, spreading the cost of Ethereum settlement across a much larger volume of activity.

Optimistic rollups, including networks based on the Arbitrum and Optimism architectures, generally assume submitted transactions are valid unless challenged during a defined period. Zero-knowledge rollups instead generate cryptographic proofs intended to demonstrate that the resulting state was calculated correctly.

The distinction remains technically important, but the competitive market is becoming less binary. Optimistic systems are improving their proof mechanisms, while zero-knowledge projects are working to reduce proving costs and improve compatibility with Ethereum applications. Both are evolving towards broader platforms capable of supporting decentralised finance, payments, gaming, consumer applications and tokenised assets.

Ethereum has also redesigned its own infrastructure around rollups. The Dencun upgrade in March 2024 introduced “blobs”, a cheaper form of temporary data storage created specifically for rollup submissions. Pectra, activated in May 2025, doubled the target number of blobs per block from three to six and raised the maximum from six to nine.

Fusaka followed in December 2025, introducing PeerDAS, or peer-to-peer data availability sampling. Instead of requiring each validator to download all rollup data, the system allows validators to sample portions of it, making it possible to increase data capacity without imposing the full burden on every node.

The result is a structural reduction in the cost of supplying Layer 2 blockspace. Rollups no longer compete directly with ordinary Ethereum transactions for the same type of storage, while Ethereum can expand rollup capacity more flexibly than before.

Cheap transactions are no longer enough

When transactions were expensive, a Layer 2 network could attract users simply by offering lower fees. That differentiator is weakening.

Fees on established rollups have fallen sharply, and competing networks can often offer broadly similar transaction economics. At the same time, Ethereum’s own main network has become less expensive as activity has migrated to Layer 2 and its capacity has increased. By early 2026, Ethereum Foundation analysis estimated that rollups were carrying approximately 95 percent of transactions associated with the wider Ethereum system.

This changes the basis of competition. A Layer 2 must now offer more than inexpensive execution. It needs users, liquidity, distribution, reliable infrastructure and applications that justify moving assets away from Ethereum mainnet or from another established network.

Base illustrates the importance of distribution. Built using the OP Stack and supported by Coinbase, it can draw on an existing exchange, wallet, payments and institutional-customer ecosystem. Its advantage does not rest on a uniquely cheap transaction. It rests on providing an accessible route from a large regulated platform into onchain applications.

Arbitrum has developed a different proposition around deep decentralised-finance liquidity, developer adoption and configurable chains. Optimism is positioning the OP Stack as shared infrastructure for a group of interconnected networks known as the Superchain. Other projects compete through zero-knowledge proofs, privacy, specialised execution or closer integration with particular applications.

The market is therefore moving from a contest among isolated chains towards competition among technology stacks and distribution ecosystems.

Every company can have a blockchain, but not every company needs one

One of the most important Layer 2 innovations is the ability to launch a customised chain without developing an entirely new blockchain protocol.

The OP Stack provides modular software for constructing Ethereum rollups, including components for execution, settlement, data availability and block production. Arbitrum offers configurable chains that allow operators to select governance, fee models, data-availability arrangements and other technical parameters.

This has lowered the barriers to creating an application-specific network. An exchange, gaming company, financial institution or consumer platform can operate dedicated blockspace, set its own fee structure and adapt the environment to its business.

The commercial appeal is clear. A company running a successful application on a public blockchain pays transaction fees to the network operator and competes with other applications for capacity. Operating its own chain allows it to control the user experience and potentially retain sequencer revenue.

It may also support requirements that are difficult to meet on a general-purpose public network, including restricted participation, custom compliance controls, predictable capacity or the use of a proprietary token to pay fees.

The risk is that launching a chain can become an infrastructure project without a compelling economic purpose. The operator assumes responsibility for upgrades, monitoring, sequencer availability, security processes, bridge management and user support. It must also persuade users and liquidity providers to enter another environment.

A dedicated Layer 2 makes sense when an organisation already has meaningful distribution, requires control over execution or can generate sufficient activity to justify operating a network. It is less convincing when a project launches a chain primarily to issue a token or imitate competitors.

Fragmented liquidity is the new scaling bottleneck

Blockchains become more useful when assets and applications can interact within the same environment. A stablecoin deposited as collateral can support a loan, be exchanged on a decentralised market and settle another transaction without leaving the chain.

Multiple Layer 2 networks divide that liquidity. An asset held on Base is not automatically available to an application on Arbitrum, even when both ultimately settle through Ethereum. Moving it may require a bridge, a separate version of the token or a third-party liquidity provider.

This creates operational friction for users and balance-sheet inefficiency for financial applications. Liquidity must be maintained across several networks, while prices and borrowing conditions can diverge. Developers must decide which chains to support, and users must understand where their assets reside.

The interface may hide some of this complexity, but the underlying risks remain. A transaction that appears to move a token instantly from one network to another may rely on an intermediary advancing liquidity before the canonical settlement process is complete.

Interoperability is consequently becoming the next major area of Layer 2 development. The Optimism Superchain is designed around networks sharing a common software stack and moving towards native communication. ZK-based ecosystems are developing groups of connected chains intended to exchange messages and assets more directly. Wallets and intent-based systems increasingly aim to let users specify an outcome without manually choosing bridges or transaction routes.

Success would make multiple chains feel like one system. Failure would leave users dependent on an increasingly complicated layer of bridging services.

Bridges remain a concentrated source of risk

A rollup may inherit important security properties from Ethereum, but that does not mean every component surrounding it is equally secure.

Bridges are particularly sensitive because they hold or control assets while representing them on another network. If a bridge contract, validator set or message-verification mechanism fails, the tokens circulating on the destination chain may no longer be fully backed.

The phrase “secured by Ethereum” can therefore be too broad. Buyers and investors need to distinguish between transaction data being published to Ethereum, execution being independently verifiable and assets being subject to upgrade keys or emergency controls held by a relatively small group.

Many rollups also continue to use centralised sequencers. The sequencer orders transactions and provides rapid confirmations, giving users a responsive experience. A sequencer outage may not allow the operator to steal funds, but it can interrupt normal activity. Transaction-ordering power may also create economic advantages through maximal extractable value.

Progress towards decentralisation is uneven. Some systems have implemented permissionless proof mechanisms or stronger constraints on administrative intervention; others retain security councils, upgrade authorities and operational controls capable of changing important parts of the system.

These controls can be useful during an emergency, particularly while technology is still developing. They also mean that Layer 2 risk cannot be assessed by Ethereum’s security record alone.

Zero-knowledge technology is becoming infrastructure

Zero-knowledge proofs were initially associated with private transactions, but their larger commercial role may be in verifying computation.

A zero-knowledge rollup can perform transactions outside Ethereum and submit a mathematical proof that the result follows the network’s rules. Ethereum verifies the proof rather than repeating every transaction. This can provide relatively rapid finality and reduce the need for a lengthy challenge period.

The technology is computationally demanding. Proof generation requires specialised software and, in some cases, substantial hardware resources. Compatibility with existing Ethereum smart contracts has also required significant engineering.

Those constraints are gradually easing. Hardware acceleration, improved proving systems and more efficient virtual machines are making zero-knowledge infrastructure practical for a wider range of uses. The same technology could eventually support verifiable offchain calculations, identity systems and privacy-preserving institutional applications.

For businesses, the meaningful development is not that one proof system will defeat every optimistic rollup. It is that cryptographic verification is becoming a purchasable infrastructure service. Companies may use proofs without building their own cryptographic teams, much as cloud customers use complex computing systems without operating data centres.

Layer 2 has a business-model problem as well as a technical one

A rollup typically earns revenue from the fees users pay and incurs costs when it publishes data or settles transactions on Ethereum. The difference can accrue to the network operator, ecosystem treasury or token-governance system.

Cheap data availability has improved rollup margins, but competition is pushing user fees down. This creates a tension between adoption and value capture. A network may process a large number of inexpensive transactions without producing substantial revenue.

Tokens add another layer of ambiguity. Some Layer 2 tokens provide governance rights but no direct contractual claim on sequencer income or network cash flow. Growth in transactions does not therefore translate automatically into value for token holders.

Optimism has sought to connect OP more closely to Superchain economics, including a 2026 proposal to use part of incoming Superchain revenue for token purchases. Other ecosystems are experimenting with staking, fee sharing and governance incentives.

Investors should treat each arrangement separately. A Layer 2 may be technologically successful while its token performs poorly, particularly when the token’s economic rights are weak or supply incentives create persistent selling pressure.

What institutions should examine before choosing a Layer 2

For a business deploying financial products, tokenised assets or payment infrastructure, the correct network cannot be selected through transaction cost alone.

Liquidity comes first. The organisation must determine whether the relevant stablecoins, collateral, trading venues and counterparties are present on the network. Low fees provide little benefit if every important transaction requires an additional bridge.

The security model should then be examined in operational terms. Who runs the sequencer? Can users force transactions through Ethereum if it becomes unavailable? Who can upgrade the contracts? How long does a canonical withdrawal take, and what external systems are used to make it faster?

Data availability is another critical distinction. Some chains publish full transaction data to Ethereum, while others use separate committees or external networks to lower costs. The latter may still be commercially appropriate, but they introduce assumptions beyond Ethereum’s validator network.

Institutions should also consider the durability of the ecosystem. A network may attract temporary activity through token incentives, only to lose users when those rewards end. Developer support, stablecoin depth, infrastructure providers and credible long-term governance are better indicators than a short period of elevated transaction counts.

Finally, an exit route matters. The organisation must know whether contracts, assets and user records can be migrated if the network becomes commercially irrelevant, changes its rules or suffers a prolonged failure.

Layer 2 innovation has solved much of the problem it was originally designed to address. Ethereum can now support far more activity at far lower cost without forcing every transaction onto its main execution layer.

The more difficult phase begins after that success. Networks must make assets portable without making bridges fragile, decentralise operations without sacrificing performance and build sustainable businesses without turning every application into another isolated chain.

Layer 2 has given Ethereum scale. Its next test is whether scale can be organised into infrastructure that users and institutions experience as one dependable market rather than dozens of inexpensive but disconnected ones.