The supplied source lists S3 storage at $0.150 per GB-month in March 2006, falling through several reductions to $0.023 in December 2016. It further states that the price is still $0.023 per GB-month today. Those observations describe the standard listed storage rate; they do not establish total workload cost, which also depends on storage class, retrieval patterns, requests, transfer, and architecture.
The mechanism is straightforward: when a mature foundational service stops cutting its headline unit price, efficiency gains for data-heavy systems must increasingly come from workload design rather than waiting for a lower baseline. The timing matters because teams are making long-horizon data and AI infrastructure commitments while object-storage footprints grow. The first implication is that retention, tiering, and access-pattern design deserve explicit economic ownership. A second implication is that cost forecasts should separately model storage, operations, and data movement. Evidence boundary: this is one author’s reconstruction of a public pricing history, not an AWS announcement or a complete cost study. Watch whether AWS changes the base S3 Standard rate or whether cost improvements continue to arrive mainly through adjacent storage classes and service features.