Work out cost per gb processed instantly with clear inputs, formula shown and shareable results.
Cost per gigabyte processed is the unit metric for data platforms, and tracking it separates growth from inefficiency. If volume grows 12 percent and cost grows 12 percent, efficiency is flat and the platform is scaling linearly — which is acceptable but means no economies of scale are being realised from batching, compression or better partitioning.
Unit cost
cost per GB = monthly cost / GB processed; cost at target = GB x target rate; linear projection = cost x (1 + volume growth)
Pick one and be consistent. Raw bytes is a better business measure since it tracks what the business generates; compressed bytes better reflects what you pay to store.
Fixed platform overhead is amortised over more volume, and larger batches amortise per-job startup cost. If it does not fall, the pipeline is dominated by genuinely variable per-byte work.