FinOps for Public Cloud

The granularity available is unmatched: billing goes down to the resource, to the hour, even to the second. That richness is also the difficulty, a massive volume of lines, continuous arrival of new services and SKUs imposing permanent maintenance of ingestion pipelines. Major specificity: Public Cloud is the category where adoption of the FOCUS specification is the most advanced, with all major providers now producing aligned exports, which makes cross-provider analysis genuinely workable, unlike SaaS or data platforms.

On allocation, native dimensions (accounts, subscriptions, projects, resource groups, tags) give direct control that few other categories offer. But the persistent difficulty is the same everywhere: shared services, centralised infrastructure and inconsistent tagging produce a share of unattributed spend that no native dimension resolves alone.

The Public Cloud trajectory has long rested on an implicit assumption: continuous organic growth in consumption, year after year, on the same hyperscalers. Regulatory and sovereignty requirements (public sector, finance, healthcare, defence) are reshaping that assumption. A share of workloads is now destined to migrate to sovereign Cloud solutions, breaking with the logic of organic growth.

This shift weakens a balance of power that has so far favoured US providers. In the context of commitment negotiations, Cloud providers regularly demand significant consumption increases, including from the largest accounts, in exchange for advantageous rate conditions. Yet those same large accounts are increasingly unable to commit to a growth trajectory with US Cloud providers, precisely because the emergence of sovereign Cloud alternatives opens up the possibility of a future reallocation of part of their workloads.

This shift has direct consequences for FinOps governance: rate commitments (Reserved Instances, Savings Plans, Committed Use Discounts) negotiated on the basis of projected continuous consumption growth lose their relevance once a significant scope moves platform. Organisations that have built their discounts on tenure and cumulative volume since their initial deployments expose themselves to a discount loss that is hard to anticipate if this migration dynamic is not factored in upstream of the contract negotiation strategy.

A mature FinOps governance function must now integrate a scenario of partial decline or reallocation towards sovereign Cloud into its trajectory assumptions, and adapt accordingly the schedule and structure of commitments taken with hyperscalers.