A micro-transaction is a payment small enough that traditional payment infrastructure cannot process it economically – typically fractions of a cent to a few cents. On blockchains like Ethereum, transaction fees often exceed the value being transferred, making naive on-chain micro-transactions impractical. Solving this problem is one of the key challenges for pay-per-use business models, streaming payments, and machine-to-machine commerce.
Why Micro-Transactions Matter#
Many useful payment patterns only work when individual transfers can be vanishingly small:
- Streaming payments – paying for media, compute, or bandwidth by the second rather than by subscription.
- Pay-per-use APIs – charging per request instead of requiring upfront commitments.
- IoT and machine-to-machine payments – autonomous devices settling resource consumption in real time.
- Content monetization – paying fractions of a cent per page view or article read, replacing advertising as a revenue model.
These patterns require transaction costs to be negligible relative to the payment amount. When a single Ethereum mainnet transaction costs several dollars in gas, sending $0.001 directly on-chain makes no economic sense.
The Gas Cost Problem#
Every transaction on a blockchain competes for block space and must pay a fee to validators. On Ethereum mainnet, even a simple ERC-20 transfer costs tens of thousands of gas units. During periods of high demand, this translates to fees that dwarf micro-transaction values by orders of magnitude.
The problem is not unique to Ethereum – any chain with meaningful decentralization faces a version of this tradeoff between security and per-transaction cost.
Approaches to Viable Micro-Transactions#
Layer 2 Rollups#
Rollups (optimistic and zero-knowledge) batch many transactions into a single proof or data submission on the base layer. This amortizes the cost of L1 security across hundreds or thousands of individual transfers, reducing per-transaction fees to fractions of a cent. Rollups inherit the security guarantees of the underlying blockchain while making micro-transactions economically viable.
State Channels#
State channels allow two parties to transact off-chain an arbitrary number of times, only settling the net result on-chain. The classic example is a payment channel: two participants lock funds in a smart contract, exchange signed balance updates off-chain, and close the channel with a single on-chain transaction. This model is well-suited to high-frequency, low-value transfers between known counterparties – such as streaming payments for a service.
The Lightning Network on Bitcoin and Raiden on Ethereum are prominent payment-channel implementations.
Application-Specific Approaches#
Some protocols sidestep the problem by aggregating micro-transactions at the application layer. A service might track usage off-chain and settle periodically – say, once a day or when a balance threshold is reached. This trades immediacy for cost efficiency and introduces a degree of trust, but it can be a pragmatic choice when the parties have an ongoing relationship.
Tradeoffs#
No solution eliminates all friction. Rollups add finality delays and depend on the health of a sequencer. State channels require both parties to be online (or delegate to a watchtower) and are poorly suited to one-off payments with strangers. Application-layer aggregation reintroduces custodial risk. The right approach depends on the use case – a streaming music service has different requirements than an IoT sensor mesh.