Metrics question
You are a PM at Google Cloud and you realized that for a set of users, VMs (Virtual Machines) have been underutilized or not utilized at all. How do you identify the root cause? How do you solve the issue and ensure that it won't happen again?
- Metrics
- Hard
Practice this question out loud. An AI interviewer asks it, follows up like a real interviewer would, and scores your answer. Type or speak.
Start a mock interview on this question · Mock interview from a job description
What this question tests
Root cause analysis and preventive systems thinking for an infrastructure product with a resource waste problem.
How to approach it
- Clarify what underutilized means, VMs running below a usage threshold, or fully idle but still provisioned and billed.
- Segment affected users, e.g. new users who over-provisioned during setup versus long-time users whose workloads changed.
- Investigate likely causes, poor default sizing recommendations, lack of usage visibility, or forgotten test environments never torn down.
- Validate with data, correlate underutilization with account age, VM creation method, and whether usage alerts exist today.
- Propose fixes, better sizing recommendations at creation time, proactive usage alerts, and auto-suggested downsizing or auto-shutdown for idle VMs.
- Define prevention metrics, percent of new VMs within recommended sizing at launch and reduction in idle VM hours quarter over quarter.
What a strong answer includes
- Separates true root cause from symptom, distinguishing bad initial sizing from forgotten idle resources as two different problems needing different fixes.
- Proposes a concrete instrumentation step, usage-based alerts, before jumping to solutions, showing a data-driven approach.
- Balances user autonomy with proactive help, e.g. suggesting rather than forcing downsizing, since some idle capacity may be intentional headroom.
- Sets a clear prevention metric, percent of new VM launches matching workload needs, to show the fix actually worked.
Common mistakes
- Jumping straight to 'auto-shutdown idle VMs' without investigating why they became idle.
- Not considering that some underutilization may be intentional, e.g. reserved headroom.
- Proposing no way to measure whether the fix actually reduced waste.
Likely follow-up questions
- How would you avoid annoying users with too many alerts.
- What would you do if auto-downsizing broke someone's production workload.
- How would you measure long term whether this recurs.
More metrics questions
- How do you define success for Yelp reviews?Google · Metrics · Medium
- Utilization went down by 45% on app XYZ in Italy for the month of August. Give a reason why and draft a plan to fix it.Spotify · Metrics · Medium
- You launched a new signup flow to encourage new users to add more profile information. A/B test results indicate that the % of people that added more information increased by 8%. However, 7 day retention decreased by 2%. What do you do?Google · Metrics · Hard
- Define the metrics for YouTube search.Google · Metrics · Medium
- You walk in to your office and find that Google cloud subscription has gone down by 20%. What do you do as a product manager?Google · Metrics · Medium
- What would be the top six metrics for WhatsApp? Now pick one from the six. Tell me three things you would do to improve that metric.Google · Metrics · Medium
More questions from Google
Learn the skill behind it
Chapters of the AI PM course that teach what this question tests.
- Chapter 9: Prove it paid off: outcomes, economics, and pricing
- Chapter 2: Data fluency: SQL, logs, and reading the truth yourself
- Chapter 14: Get the job: the AI PM interview loop