GCP PCA Study Guide 2026: How to Pass Google Cloud Professional Cloud Architect

The Google Cloud Professional Cloud Architect (PCA) is the expert-tier architecture certification on GCP — one step up from Associate Cloud Engineer (ACE). Where ACE tests whether you can operate and deploy on Google Cloud, PCA tests whether you can design the solution in the first place: translating business requirements into an architecture, defending trade-offs, and reasoning through security, cost, and reliability all at once. The exam is built entirely around four published case studies, which makes PCA prep meaningfully different from a normal domain-by-domain study plan. This guide covers both.

GCP PCA Exam at a Glance

There is no formal prerequisite to sit PCA, but Google recommends 3+ years of industry experience with at least 1 year designing solutions on Google Cloud. Unlike ACE, PCA does not have a fixed passing percentage published — budget your prep time around understanding the case studies deeply rather than memorizing a target score. The certification is valid for 2 years.

The 6 GCP PCA Domains You Must Master

Designing and Planning (Domain 1) carries the highest weight at 24% — a clear signal that PCA rewards architectural judgment over hands-on execution. Compare this to ACE, where Deploying and Implementing (the "how") carries the most weight. PCA questions constantly ask you to justify the "why" behind a design choice against a specific case study's stated constraints.

Domain 1: Designing and Planning a Cloud Solution Architecture (24%)

• ●Translating business requirements (budget, SLAs, compliance, time-to-market) into concrete architecture decisions
• ●High availability and disaster recovery design: multi-zone vs multi-region, RTO/RPO trade-offs stated in the case study
• ●Migration strategy selection: lift-and-shift vs replatform vs refactor, and when to use Migrate for Compute Engine or Database Migration Service
• ●Designing for horizontal scalability: stateless service design, load balancer tier selection, autoscaling policy design
• ●Architecture-level cost optimization: committed use discounts, sustained use discounts, right-sizing recommendations
• ●Choosing managed vs self-managed services based on the case study's stated operational maturity, not just technical fit

Domain 2: Managing and Provisioning a Solution Infrastructure (15%)

• ●Hybrid connectivity design: Cloud VPN vs Dedicated Interconnect vs Partner Interconnect, chosen by bandwidth and latency requirements
• ●Shared VPC and resource hierarchy design for multi-team, multi-project organizations
• ●Infrastructure as Code: Terraform and Deployment Manager for repeatable, auditable provisioning
• ●Compute provisioning trade-offs at scale: Compute Engine vs GKE vs serverless, matched to team operational maturity
• ●Data pipeline provisioning for the case-study workloads: Pub/Sub, Dataflow, and BigQuery working together

Domain 3: Designing for Security and Compliance (18%)

• ●IAM design at organization scale: custom roles, Organization Policy constraints as compliance guardrails
• ●Data protection: encryption at rest and in transit, Cloud KMS, and choosing CMEK vs Google-managed keys
• ●Regulatory compliance mapping — e.g. HIPAA considerations for the EHR Healthcare case study
• ●Zero-trust access patterns: VPC Service Controls, Identity-Aware Proxy, and private Google access
• ●Auditing and visibility: Cloud Audit Logs and Security Command Center

Domain 4: Analyzing and Optimizing Technical and Business Processes (18%)

• ●Defining technical processes: CI/CD pipeline design, release management, and rollback strategy
• ●Defining business processes: SLA/SLO definition and disaster recovery runbooks
• ●DevOps culture and incident management: on-call structure, postmortems, and continuous improvement loops
• ●Confirming the solution actually meets stated business objectives through ongoing cost and performance monitoring

Domain 5: Managing Implementation (11%)

• ●Advising development teams on implementation approach without writing the application code yourself
• ●Interacting with Google Cloud programmatically: client libraries and gcloud SDK integrated into CI/CD pipelines
• ●Overseeing end-to-end implementation of the case study's data and compute architecture against the original design

Domain 6: Ensuring Solution and Operations Reliability (14%)

• ●Designing monitoring and alerting strategy: Cloud Monitoring dashboards and SLO-based alerting policies
• ●Deployment and release patterns: blue/green and canary rollouts for zero-downtime releases
• ●Support model design: on-call rotation structure and escalation paths
• ●Troubleshooting strategy and structured post-incident review processes

The 4 PCA Case Studies You Must Know

Google publishes four case studies that the entire PCA exam is built around. Several questions reference a case study's specific stated constraints — budget limits, compliance requirements, existing infrastructure — so read all four in full before your exam date, not just a summary.

• ●TerramEarth — a heavy equipment manufacturer moving from batch IoT telemetry uploads to real-time streaming analytics, with strict data residency and offline-vehicle sync requirements
• ●Helicopter Racing League — a live-streaming and machine-learning company needing global low-latency video distribution alongside real-time race analytics
• ●Mountkirk Games — a gaming studio building a real-time multiplayer backend that must scale globally with leaderboard and matchmaking services
• ●EHR Healthcare — a regulated healthcare SaaS provider migrating to GCP under HIPAA compliance, with strict uptime and disaster recovery requirements

GCP PCA Study Plan: 8-Week Schedule

1. 1Week 1: Read all four official case studies twice — once for business context, once for the specific technical constraints each one states
2. 2Week 2: Domain 1 — architecture design patterns, HA/DR trade-offs, migration strategy selection against each case study
3. 3Week 3: Domain 2 — hybrid connectivity, Shared VPC design, Terraform/Deployment Manager provisioning
4. 4Week 4: Domain 3 — IAM at organization scale, Cloud KMS, VPC Service Controls, and HIPAA-style compliance mapping (EHR Healthcare)
5. 5Week 5: Domain 4 — SLA/SLO definition, CI/CD process design, incident management processes
6. 6Week 6: Domain 5 and 6 — implementation oversight, monitoring/alerting design, deployment patterns
7. 7Week 7: Re-read each case study and draft your own architecture diagram + trade-off justification for each — this is the single highest-value PCA prep exercise
8. 8Week 8: Full mock exam practice (2 full exams) and review every missed question against the relevant case study constraint

Common GCP PCA Exam Traps

• ●The case study's stated constraint always overrides generic best practice — if a case study specifies a fixed budget or existing on-prem investment, the "textbook best" answer may be wrong for that scenario
• ●PCA questions often have two technically correct answers — the right one is the one that fits the specific case study's stated priorities (cost vs latency vs compliance)
• ●Don't confuse "managed" with "simplest" — a case study emphasizing full control or existing team expertise may point toward Compute Engine over a fully managed service
• ●HIPAA and other compliance requirements in EHR Healthcare are not optional add-ons — treat them as hard constraints that eliminate otherwise-valid architecture options
• ●Multi-region does not automatically mean "more available" if the case study's actual bottleneck is a single-region dependency elsewhere in the design (e.g. a regional database)
• ●Cost optimization questions expect committed-use or sustained-use discounts as the answer for steady, predictable workloads — not just "use a cheaper VM"

Best Study Resources

• ●Google's official case study PDFs — read the full text, not summaries; several exam questions quote details directly
• ●Official PCA exam guide — published by Google, maps every topic to its domain
• ●Google Cloud Architecture Center — reference architectures and design pattern whitepapers that mirror PCA's reasoning style
• ●Google Cloud Skills Boost — hands-on labs to ground the theory, especially for Domain 2 provisioning tasks
• ●CloudTechTrainings live batch — instructor-led sessions covering both ACE and PCA domains together

Exam Day Strategy

PCA gives you roughly 2.2 minutes per question, but several questions require re-reading a section of a case study before you can answer confidently — budget extra time for that rather than treating it like a flat per-question limit.

• ●Skim the relevant case study section again before answering, rather than relying on memory from a week of study
• ●When two options both seem technically valid, pick the one that matches the case study's explicitly stated priority (cost, compliance, latency, or existing skillset)
• ●Flag long scenario questions and come back to them after clearing shorter, more direct questions first
• ●Watch for questions that test whether you can say no to over-engineering — the "right" PCA answer is sometimes the simpler design that still meets every stated requirement
• ●If a question conflicts with something you know is generally true on GCP, re-check whether the case study states a constraint that changes the answer

ACE vs PCA — Which Should You Take First?

ACE is not an official prerequisite for PCA, but it is the natural first step if you are newer to GCP — ACE builds the hands-on console and CLI muscle memory that PCA then tests you on at a design level. If you already have solid GCP production experience, you can go straight to PCA. See our GCP ACE study guide for the associate-level foundation, or jump straight into PCA prep if you're already comfortable deploying on GCP.

Ready to Start Your Cloud Journey?

Live batches Mon–Sat — Azure 9–10 AM IST (starts 3 august 2026) · AWS 10:30–11:45 AM IST (started 6 july 2026 — join in progress). Hands-on labs, exam prep, and community support.

Keep Reading

GCP

GCP ACE Study Guide 2026: How to Pass Google Cloud Associate Cloud Engineer

A practical study plan for the Google Cloud Associate Cloud Engineer exam. Covers all 5 domains with key service mappings, common traps, and exam-day tips to pass ACE first attempt.

8 Jul 2026·11 min read

Read →

Azure

AZ-204 Is Retiring on 31 July 2026: What Azure Developers Should Do Now

Microsoft retires the AZ-204 exam, the Azure Developer Associate certification, and its renewal assessment on 31 July 2026 — replaced by the new AI-200 Azure AI Cloud Developer Associate. Here is exactly what changes, what does not, and what to do depending on where you stand.

22 Jul 2026·8 min read

Read →