One of the key benefits of distributed ledger technologies is decentralized, tamper-proof data storage. A blockchain network such as Ethereum or Bitcoin can take data transmitted via transactions, add it to a block and persist copies of it on every node. As long as the network is alive the data can be recovered and used independently of the action or inaction of any individual agent including the original sender of the data, who would normally have to incur perpetual expenses for hosting and securing the data.

The network charges a one-time fee for providing this service for its remaining lifetime to cover energy, capital and labor costs over a timespan of unknown length. The fee is charged at transaction time and in the case of Ethereum is calculated as:

Gas Usage [units of gas] *

Gas Cost [Ether / gas] *

Network Token Price [$ / Ether]

Transaction gas usage is a function of bandwidth, storage and compute resources consumed by the transaction and can be calculated before submitting the transaction. It is a nominal amount that does not represent its real resource cost. An identical transaction will always use the same amount of gas regardless of the state of the network at the time it is submitted unless the network protocol itself is changed between transactions.

Gas cost is chosen by the user. The higher the amount offered by the user the faster the transaction will be included in a block. The relationship can be estimated from network statistics.

The fiat currency price of the network token (in this case Ether) is a market price that is determined on exchanges where network tokens are traded for fiat currency.

All three factors combined should (but not necessarily do) represent the real resource cost incurred by the transaction discounted to the present. The resulting value is transferred from the user to the network (miners in the case of Ethereum and Bitcoin).

It is reasonable to expect that networks will compete for users based on this resource cost, i.e. if it is cheaper to store data of a certain size on network A compared to an otherwise identical network B then network A should acquire more users interested in this particular feature, for example software developers who require self-updating local storage for building temporarily disconnected applications.