What does the Google Cloud Professional Cloud Architect exam test?
The Professional Cloud Architect (PCA) exam tests your ability to design, develop, and manage robust, secure, scalable, and dynamic solutions to drive business objectives using Google Cloud technologies. Unlike associate-level exams, PCA heavily emphasizes architectural decision-making through detailed case studies and requires justifying trade-offs between competing design approaches.
The Google Cloud Professional Cloud Architect certification is consistently ranked among the most valuable cloud credentials in the industry. In Global Knowledge's 2025 IT Skills and Salary Report, PCA-certified architects reported median annual salaries above $170,000 in North America, placing it in the top five cloud certifications by compensation globally [1]. The exam demands more than service memorization; it requires genuine architectural judgment, the ability to read case studies, and the skill to defend design decisions against multiple plausible alternatives.
This guide covers the full exam structure, all case studies, domain-by-domain content breakdown, and a preparation strategy validated by current exam candidates. Sources include Google's official PCA exam guide [2], the Google Cloud Architecture Center [3], Linux Foundation's 2025 open source skills report [4], and Coursera's cloud career survey data [5].
Exam Overview
| Attribute | Detail |
|---|---|
| Exam cost | $200 USD |
| Exam duration | 120 minutes |
| Number of questions | 50-60 multiple-choice and multiple-select |
| Case studies | 4 pre-published case studies; questions reference these scenarios |
| Validity period | 2 years |
| Delivery | Remote proctored (Kryterion) or test center |
| Prerequisites | None (Google recommends 3+ years IT experience, 1+ year GCP) |
| Recertification | Pass full exam again; no delta exam option |
The $200 registration fee is higher than associate-level exams. Many employers reimburse this cost when candidates pass; verify your employer's tuition assistance policy before self-funding.
The Case Study Format
PCA is unusual among cloud certifications in that Google pre-publishes four case studies which appear in the actual exam. Candidates who have studied these case studies thoroughly answer scenario questions dramatically faster, leaving more time for novel architectural reasoning questions.
The four current case studies are:
Cymbal Bank -- A financial services institution migrating from legacy on-premises systems to GCP, requiring compliance with financial industry regulations, data residency requirements, and zero-downtime migration strategies.
Dress4Win -- A social fashion platform scaling from a single data center to GCP, emphasizing containerization, database migration, and maintaining high availability during transition.
Mountkirk Games -- An online gaming company building a globally distributed multiplayer game backend, emphasizing low-latency compute, global load balancing, and real-time analytics.
TerramEarth -- A heavy equipment manufacturer collecting telemetry data from machinery in the field, emphasizing IoT data ingestion, streaming analytics, and predictive maintenance architectures.
"Candidates should read all four case studies multiple times before the exam and practice justifying specific GCP service choices against the business and technical requirements stated in each scenario." -- Google Cloud certification community guidance [2]
Memorize the key requirements, constraints, and existing infrastructure for each company. Exam questions will name the company and assume you recall the details without re-presenting the full case study text.
Exam Domains
Google's official exam guide divides PCA content into six domains:
| Domain | Title | Approximate Weight |
|---|---|---|
| 1 | Designing and planning a cloud solution architecture | 24% |
| 2 | Managing and provisioning a solution infrastructure | 15% |
| 3 | Designing for security and compliance | 18% |
| 4 | Analyzing and optimizing technical and business processes | 16% |
| 5 | Managing implementation | 11% |
| 6 | Ensuring solution and operations reliability | 16% |
Domain 1: Designing and Planning a Cloud Solution Architecture (24%)
The highest-weighted domain requires you to translate business requirements into GCP architectures. Questions present a business scenario and ask you to select the correct set of GCP services or reject inadequate designs.
Key Sub-Topics
Compute architecture selection: When to use Compute Engine (long-running, stateful workloads with fine-grained control), GKE (containerized workloads requiring orchestration), App Engine (managed runtime web applications), Cloud Run (stateless container workloads with variable traffic), and Cloud Functions (event-driven, single-purpose functions).
High availability and disaster recovery: Understanding RPO (Recovery Point Objective) and RTO (Recovery Time Objective) targets and mapping them to GCP deployment patterns. A 99.99% availability requirement implies multi-region active-active; a 99.9% requirement may permit regional failover.
Network architecture: Designing VPC topologies including hub-and-spoke with Shared VPC, Cloud Interconnect vs. Cloud VPN trade-offs, and Private Google Access for internal workloads.
Migration strategy: Choosing between lift-and-shift (Migrate to Compute Engine), re-platforming (containerization), and refactoring (serverless or managed services) based on time-to-production, cost, and risk tolerance.
Domain 2: Managing and Provisioning a Solution Infrastructure (15%)
This domain tests infrastructure lifecycle management rather than initial architecture.
Infrastructure as Code
- Terraform for GCP: resource blocks for all major GCP services, state management with Cloud Storage backends, and workspace-based environment separation
- Cloud Deployment Manager: YAML-based templates, still present in exam but Terraform is increasingly preferred in scenario answers
- Config Connector: Kubernetes-native resource management for GCP services
Resource Hierarchy Management
- Organization, folder, and project structure aligned to business units and compliance boundaries
- IAM policy inheritance and override patterns across the hierarchy
- Shared VPC architecture: host project, service projects, and network admin roles
Domain 3: Designing for Security and Compliance (18%)
Security architecture is heavily tested on PCA. This domain overlaps with the Professional Cloud Security Engineer exam but is tested at a design level rather than configuration level.
Identity and Access Design
- Designing IAM structures for large organizations with dozens of teams
- Workforce Identity Federation for connecting enterprise identity providers (Active Directory, Okta) to GCP without managing individual user accounts
- Workload Identity Federation replacing service account keys for GKE and CI/CD workloads
Data Protection Architecture
- Customer-managed encryption keys (CMEK) with Cloud KMS: key rings, key versions, and rotation schedules
- VPC Service Controls: defining access policies that prevent data exfiltration even if IAM is misconfigured
- Data Loss Prevention (Cloud DLP): de-identification techniques for PII in BigQuery and Cloud Storage
Compliance Requirements
- Understanding how GCP features map to regulatory frameworks: HIPAA, PCI DSS, SOC 2, FedRAMP
- Organization Policy constraints for enforcing compliance guardrails: restricting resource locations, disabling public IP on VMs, requiring OS Login
"Security design on the PCA exam requires understanding both what controls exist and when to apply them. The exam tests judgment about proportionate security measures, not just knowledge of available features." -- Google Cloud Architecture Center [3]
Domain 4: Analyzing and Optimizing Technical and Business Processes (16%)
This domain is the most abstract, testing your ability to recommend process improvements alongside technical solutions.
Cost Optimization
- Committed Use Discounts (CUDs) vs. Sustained Use Discounts (SUDs): CUDs require 1-3 year commitments for 37-55% discounts; SUDs apply automatically for compute usage exceeding 25% of the month
- Rightsizing VMs using the Recommender API
- Cloud Storage lifecycle policies to automatically transition objects to lower-cost classes
- BigQuery reservations and slot commitments vs. on-demand pricing
CI/CD and DevOps Integration
- Cloud Build for automated testing and deployment pipelines
- Artifact Registry for container image and package management
- Cloud Deploy for managed continuous delivery to GKE and Cloud Run
Business Process Analysis
- Translating qualitative business requirements (reduce operational overhead, improve time-to-market) into specific GCP service recommendations
- Evaluating total cost of ownership for managed services vs. self-managed alternatives
Domain 5: Managing Implementation (11%)
This lighter-weighted domain covers the handoff between architecture and engineering teams.
- Writing and reviewing implementation guides for development teams
- Advising development teams on application design for GCP: connection pooling for Cloud SQL, exponential backoff for API calls, idempotent operations for Pub/Sub consumers
- Integration testing strategies for GCP workloads: using Pub/Sub emulator, Firestore emulator, and Cloud Storage local equivalents
Domain 6: Ensuring Solution and Operations Reliability (16%)
Reliability engineering concepts from Site Reliability Engineering (SRE) practice appear throughout this domain.
Observability Architecture
- Designing a complete observability stack: Cloud Monitoring for metrics, Cloud Logging for structured logs, Cloud Trace for distributed tracing, Cloud Profiler for application profiling
- Defining SLIs (Service Level Indicators) and SLOs (Service Level Objectives) for GCP-hosted applications
- Creating alerting policies that fire on symptom-based signals rather than cause-based signals
Operational Procedures
- Designing runbooks for common failure scenarios
- Configuring managed instance group health checks and autohealing policies
- Backup and restore procedures for Cloud SQL, Cloud Spanner, and Cloud Storage
Preparation Strategy
Phase 1: Service Breadth (Weeks 1-4)
The PCA exam requires familiarity with over 80 GCP services. You cannot master all of them, but you must understand the primary use case, key trade-offs, and pricing model for each major service family.
Prioritize in this order:
- Compute (Compute Engine, GKE, Cloud Run, App Engine, Cloud Functions)
- Storage and Databases (Cloud Storage, Cloud SQL, Cloud Spanner, Firestore, Bigtable, Memorystore)
- Networking (VPC, Load Balancing, Cloud CDN, Cloud Interconnect, Cloud Armor)
- Security (IAM, Cloud KMS, VPC Service Controls, Security Command Center)
- Data and Analytics (BigQuery, Pub/Sub, Dataflow, Dataproc)
- DevOps (Cloud Build, Artifact Registry, Cloud Deploy, Terraform)
Phase 2: Case Study Mastery (Weeks 5-6)
Download all four case studies from the Google certification page. For each one:
- List every stated business requirement
- List every stated technical requirement
- List every constraint (compliance, existing systems, budget)
- Map each requirement to specific GCP services
- Identify which requirements create tension with each other
Practice answering: "Which services would you recommend for Cymbal Bank's data warehouse and why?" without looking at your notes.
Phase 3: Practice Exams and Architectural Reasoning (Weeks 7-8)
| Resource | Type | Recommended |
|---|---|---|
| Tutorials Dojo PCA Practice Exams | Practice questions | Yes |
| Whizlabs PCA Practice Tests | Practice questions | Yes |
| Dan Sullivan "Professional Cloud Architect Study Guide" | Textbook | Yes |
| Google Cloud Architecture Center | Reference docs | Yes |
| A Cloud Guru PCA Course | Video course | Optional |
After each practice exam, spend at least as much time reviewing incorrect answers as you spent taking the test. For every wrong answer, identify whether the error was: service confusion, case study detail recall, architectural judgment, or CLI/API specifics.
"The PCA exam rewards candidates who can explain why a service is wrong for a scenario, not just which service is right. Understanding the failure modes and limitations of GCP services is as important as knowing their features." -- Tutorials Dojo PCA exam analysis [6]
Distinguishing Architecture Patterns
Several recurring architectural patterns appear across PCA exam scenarios. Recognizing these patterns enables faster and more accurate answer selection.
Event-Driven Architecture: Use Pub/Sub as the message bus, Cloud Functions or Cloud Run for event consumers, and Firestore or BigQuery as the downstream store. Apply when decoupling producers from consumers is required.
Batch Analytics Pipeline: Cloud Storage for raw data landing, Dataflow for transformation (Apache Beam), BigQuery for analysis, and Looker Studio for visualization. Apply when processing large volumes of historical data on a schedule.
Real-Time Streaming Pipeline: IoT Core or Pub/Sub for ingestion, Dataflow in streaming mode for transformation, BigQuery streaming inserts for storage, and Looker for dashboards. Apply for TerramEarth-style scenarios.
Microservices on GKE: GKE Autopilot for cluster management, Cloud Load Balancing for ingress, Cloud SQL with private IP for persistence, Artifact Registry for images, and Cloud Build for CI/CD.
Serverless Web Application: Cloud Run for API backend, Cloud Storage for static assets with Cloud CDN, Firestore for user data, Firebase Authentication for identity, and Secret Manager for API keys.
References
[1] Global Knowledge. "IT Skills and Salary Report 2025." globalknowledge.com. Accessed May 2026.
[2] Google Cloud. "Professional Cloud Architect Exam Guide." cloud.google.com/certification/cloud-architect. Accessed May 2026.
[3] Google Cloud. "Cloud Architecture Center." cloud.google.com/architecture. Accessed May 2026.
[4] Linux Foundation. "State of Tech Talent Report 2025." linuxfoundation.org. Accessed May 2026.
[5] Coursera. "Job Skills Report 2025." coursera.org. Accessed May 2026.
[6] Tutorials Dojo. "Google Certified Professional Cloud Architect Practice Exams." tutorialsdojo.com. Accessed May 2026.
[7] Google Cloud. "SRE Book: Site Reliability Engineering." sre.google/sre-book. Accessed May 2026.