Architecting for Value

A significant share of a technology workload’s lifecycle cost is determined by decisions made during its design and placement, even before a single resource is provisioned. The Architecting & Workload Placement capability in the FinOps Framework formalizes this, connecting architectural decisions and workload placement options to business goals through unit metrics and value comparisons.

This also includes “shifting left”, embedding financial requirements, cost estimation, and value alignment earlier in the engineering and product lifecycle, before workloads reach production or as they are being designed. The State of FinOps 2026 data shows this shift is underway: pre-deployment architecture costing emerged as a top desired tooling capability, and Platform Engineering teams are increasingly joining FinOps conversations.

Architecting for value applies across all Technology Categories, including not only public cloud resources, but SaaS platforms, data cloud environments, data center placement, and also AI workloads, where the choice between API-based inference, fine-tuned models, and self-hosted infrastructure can have significant cost and value implications.