Google Cloud Professional Cloud Architect study guide

How to use this guide

What to focus on:

• Translating business and technical requirements into an architecture
• Designing for availability, durability and failure, including regional vs multi-regional choices
• Network design: VPC layout, Shared VPC, hybrid connectivity, Private Service Connect
• Cost modelling, and choosing managed services over self-managed where it pays

What to focus on:

• Provisioning compute, storage and data services to match the workload
• Infrastructure as code and repeatable environment configuration
• Migration planning: lift-and-shift versus replatforming, and cutover strategy
• Networking implementation: routing, load balancing, DNS, NAT

What to focus on:

• IAM design: least privilege, service accounts, workload identity federation
• Data protection: CMEK, secret management, encryption in transit and at rest
• Separation of duties and organisation policy constraints across a resource hierarchy
• Meeting regulatory requirements and designing for auditability

What to focus on:

• Reliability practice: SLIs, SLOs, error budgets and what they change
• Cost optimisation: committed use, autoscaling, rightsizing, storage classes
• Incident and change management processes that hold up under load
• Team and stakeholder considerations in a technical decision

What to focus on:

• Advising development and operations teams on Google Cloud practice
• CI/CD pipeline design and safe deployment strategies (canary, blue-green)
• Interacting with the platform programmatically through APIs, CLI and client libraries
• Testing and validating an implementation against the design

What to focus on:

• Observability: logging, monitoring, tracing and alerting that fires usefully
• Post-incident review, and turning findings into architectural change
• Ongoing optimisation as usage and requirements shift
• Deprecation and lifecycle management of deployed solutions

Common mistakes

• ✕Studying services in isolation instead of practising trade-offs. The exam almost never asks what a product does, only which one fits the constraints given.
• ✕Skimming the case studies. Google publishes them in advance; going in cold wastes time you needed for the scenario questions.
• ✕Treating design and planning as one domain among six when it is 25% of the exam on its own.
• ✕Defaulting to the most capable service rather than the most appropriate. Over-engineering is a wrong answer when the requirements state a cost constraint.
• ✕Assuming a passing percentage exists and calibrating to it. Google publishes none, so practice-set consistency is the only useful gauge.

Exam-day tips

• ✓Re-read the published case studies beforehand. Questions referencing them go much faster when the context is already familiar.
• ✓Anchor on the constraint the scenario states explicitly (budget, RPO, compliance regime). It is usually what separates two otherwise-reasonable options.
• ✓Multiple-select items tell you how many to choose. Count them, because partial answers score nothing.
• ✓Two hours across 50 to 60 questions is roughly two minutes each. Flag the long scenarios and return rather than stalling.
• ✓When two options both work, prefer the managed service unless the scenario gives a specific reason not to.

How to know you’re ready

One good practice score can be luck. What counts is scoring at or above the real pass mark (Not published by Google) across several full-length sets in a row, with no single domain trailing far behind the others. Kwizza tracks your per-domain readiness automatically as you practice.