Question

A retail company has been running on Google Cloud for two years. Their FinOps team reports that Compute Engine spending has grown 40% year over year, but application performance dashboards show most production VMs consistently using under 20% CPU and 30% memory. Leadership wants a data-driven, low-risk process to reduce compute waste without impacting availability. Which approach should the cloud architect recommend to systematically identify and act on the savings opportunity?

Accepted Answer

Review the Compute Engine rightsizing recommendations in the Recommender / Active Assist console, validate them against workload metrics, and apply machine-type changes through a controlled change process.

Answer

Immediately migrate all underutilized VMs to the smallest available machine type to guarantee the maximum possible cost reduction.

Answer

Purchase three-year committed use discounts for the entire current fleet to lock in lower pricing before optimizing usage.

Answer

Convert all production VMs to Spot instances since utilization is low and the workloads clearly do not need guaranteed capacity.