Cloud cost optimization interview questions and answers

Key takeaways

• Rightsizing and autoscaling match capacity to actual demand.
• Reserved or savings-plan pricing suits steady, predictable workloads; spot suits flexible, interruptible ones.
• Storage tiering moves infrequently accessed data to cheaper tiers.
• Tagging and cost allocation make spending visible by team or project.
• Optimisation is a continuous process, not a one-time exercise.

Rightsizing

1. What is rightsizing?

Matching the size of a resource, such as a VM or database instance, to its actual usage, instead of leaving it over-provisioned "just in case."

2. How do you find over-provisioned resources?

Look at CPU, memory and I/O utilisation over a meaningful period, such as two to four weeks, and flag resources running well below capacity most of the time.

Commitment discounts and spot capacity

3. What is a reserved instance or savings plan?

A commitment to use a certain amount of compute over a period, usually one or three years, in exchange for a lower price than on-demand.

4. When should you avoid reserved capacity?

When the workload is new, its size is uncertain, or it is expected to shrink or change shape soon. Committing early can lock in waste.

5. What are spot or preemptible instances?

Spare capacity sold at a steep discount that can be reclaimed by the provider with short notice. They suit batch jobs, stateless workers and fault-tolerant workloads, not stateful production databases.

6. How would you design a workload to use spot capacity safely?

Make it stateless or checkpointed, handle interruption gracefully, and keep a fallback to on-demand capacity for critical paths.

Storage and data

7. What is storage tiering?

Moving data to cheaper storage classes as it becomes less frequently accessed, for example from hot storage to cool or archive tiers.

8. How do you reduce data transfer costs?

Keep traffic within the same region or availability zone where possible, use a content delivery network for public content, and monitor cross-region or egress traffic.

Autoscaling and architecture

9. How does autoscaling save money?

It adds capacity when demand rises and removes it when demand falls, so you are not paying for peak capacity around the clock.

10. What is the cost impact of serverless compute?

You pay per execution rather than for idle capacity, which can be cheaper for spiky or infrequent workloads and more expensive for constant, high-volume ones. Evaluate both.

Visibility and governance

11. Why is tagging important for cost optimisation?

Tags let you attribute spend to teams, projects or environments, which is the first step to knowing where to optimise.

12. What is a cost anomaly, and how do you catch one?

An unexpected spike in spend, caught with budget alerts and anomaly detection tools rather than finding it on the monthly bill.

13. How do you avoid "zombie" resources?

Set up regular reviews or automated checks for unattached storage volumes, idle load balancers, old snapshots and unused IP addresses, and remove them.

14. Who should own cost optimisation in an organisation?

It works best as a shared responsibility between engineering, finance and leadership, often called FinOps, rather than sitting with one team alone.

Practical and behavioural

15. Describe how you would optimise a workload costing more than expected.

Check rightsizing first, then look at storage tiers, idle resources, and whether reserved or spot pricing fits the usage pattern. Make changes incrementally and measure the effect.

16. What trade-offs come with aggressive cost optimisation?

Over-optimising can hurt performance, resilience or developer speed. Balance savings against the cost of outages or lost productivity.

How should you prepare?

Practise in a free-tier cloud account: launch a few resources, check their utilisation, and try rightsizing and tagging them yourself. The AWS documentation and Google Cloud documentation both explain their pricing models in detail. For the broader interview ground, see cloud cost management interview questions.

Next steps

Build hands-on cloud skills with our AWS Solutions Architect Associate course or AWS DevOps Engineer Professional course.

Related reading

• [2026] Top 50+ Cloud Scalability Interview Questions and Answers
• [2026] Top 50+ Cloud Governance Interview Questions and Answers