A retail company's most critical application is its online payment processing system.

The Google Cloud Architecture Framework: Reliability section emphasizes that for applications requiring protection against zonal failures, the most cost-effective solution is a Regional Managed Instance Group (MIG). A Regional MIG automatically distributes VM instances across multiple zones within a single region. If one zone fails, the MIG’s auto-healing and cross-zone load balancing ensure that instances in the remaining healthy zones continue to serve traffic.

Option B is the correct choice because it satisfies the "survive a zonal outage" requirement while "minimizing cost." Multi-region deployments (Option A) provide higher availability (surviving regional outages) but significantly increase costs due to inter-region data transfer and the expense of multi-region database configurations like Spanner. Spot VMs (Option C) are unsuitable for a "critical payment processing system" because they can be preempted at any time, risking transaction failures. Snapshot recovery (Option D) results in high Recovery Time Objectives (RTO), failing the performance expectations of a critical retail application. By using a Regional MIG with a regional database (like Cloud SQL with High Availability enabled), the company achieves robust zonal resilience at a much lower price point than a global multi-region architecture.