Google Cloud Professional Cloud Architect (PCA) – Study Plan

Overview

The Google Cloud Professional Cloud Architect exam validates your ability to design scalable, secure, reliable, and cost-optimized architectures on Google Cloud. Success here requires more than just memorizing service features; you have to demonstrate the ability to apply that knowledge to real-world scenarios.

This exam tests architectural judgment. The questions will place you in situations where you need to recommend the best solution, weigh the pros and cons of competing designs, and map specific GCP tools to complex business requirements. This mirrors the daily responsibilities of a Cloud Architect, and this study plan is designed specifically to build and practice that skill set.

The October 2025 update (v6.1) brought meaningful changes. AI and ML are now firmly in scope, the Well-Architected Framework is required knowledge, and "Securing AI" is an explicit topic. All of that is reflected in the weeks ahead.

Case studies: 4 available case studies; each exam sitting includes 2 (randomly selected). Case study questions make up 20-30% of the exam.

The Google Cloud Well-Architected Framework (six pillars: operational excellence, security, reliability, performance optimization, cost optimization, and sustainability) is now required knowledge. It is used as evaluation criteria across the entire exam. Internalize these pillars early. They will shape how you think about every design decision from Week 1 onward.

Recommended experience: 3+ years of industry experience including 1+ year designing and managing solutions using Google Cloud. No formal prerequisites.

A note on prerequisites: There are no formal prerequisites for PCA, and many candidates pass without holding ACE first. That said, if GCP is new to you, ACE builds the operational foundation that this exam assumes. Consider it a worthwhile detour, not a roadblock.

Frequently Asked Questions

Do I need ACE before PCA?

There are no formal prerequisites, and many people pass PCA without it. But if you have not spent significant time working with GCP at an operational level, the infrastructure management and operations domains will feel unfamiliar. ACE builds that foundation.

Why is this plan 8 weeks instead of 10 or 12?

Because you are expected to have 1+ year of hands-on GCP experience coming in. This plan skips service introductions and focuses on architectural judgment: when and why to use services, not what they are. Compute and networking are studied together because you deploy compute inside networks. Reliability and operations are combined because keeping systems running and improving them is the same discipline. Data analytics and AI/ML are combined because pipelines feed into ML workloads. Same ground, tighter structure.

What are the four case studies for 2026?

Altostrat Media, Cymbal Retail, EHR Healthcare, and KnightMotives Automotive. Each exam sitting gives you 2 of the 4 (randomly selected), and case study questions make up 20-30% of the exam. If you come across study materials referencing Dress4Win, TerramEarth, Mountkirk Games, JencoMart, or Helicopter Racing League, those are outdated.

How many practice questions should I aim for?

Aim for at least 600 unique questions across all domains, ensuring you prioritize quality over quantity. ExamOS provides over 2,000 high quality practice questions covering all domains and designed to explain the rationale behind both correct and incorrect answers.

When should I book the exam?

Once you are consistently hitting 80% or above on Legend mode across two or more sessions, and you feel comfortable explaining architecture decisions for all four case studies, you are ready. Trust your preparation.

How long is the exam?

120 minutes (2 hours) for 50-60 questions. Case study questions tend to take longer than standalone questions, so keep an eye on the clock. If you are renewing an active certification, the renewal exam is shorter: 1 hour, 25 questions, 1 case study.

What languages is the exam available in?

English and Japanese only. This is more limited than most Microsoft and AWS exams. Google does not offer additional time accommodations for non-native speakers.

What is the "best fit" heuristic?

When a managed GCP service can meet the requirements, it is almost always the right answer over a self-managed alternative. The exam rewards solutions that keep operational overhead low. When two managed services seem similar, go with the one purpose-built for the workload at hand.

Can I use reference materials during the exam?

No. The PCA exam is closed-book. You will not have access to documentation, websites, or search engines. The case studies are available on a split-screen view during the exam, but that is all the reference material you get.

What changed in the October 2025 update (v6.1)?

Several meaningful changes:

• The Well-Architected Framework went from optional best practice to required knowledge. The operational excellence pillar got its own dedicated section.
• AI and ML content expanded significantly. New sections were added for Vertex AI workflows and pre-built AI solutions. "Securing AI" became a specific security topic.
• Terraform/IaC became an explicit requirement for the infrastructure management domain.
• Case studies were refreshed. EHR Healthcare was kept. Three new scenarios (Altostrat Media, Cymbal Retail, KnightMotives Automotive) replaced the older ones. All three incorporate AI integration.
• Existing certifications stay valid until their normal expiration date.

What is the most heavily tested domain?

Designing and planning a cloud solution architecture is the largest at roughly 24%. But the exam evaluates your answers against the Well-Architected Framework across all six domains. Understanding how the six pillars apply to every design decision matters more than focusing on any single domain.

What are the most common ways people fail?

• Studying old case studies. Dress4Win, Mountkirk Games, TerramEarth, and Helicopter Racing League are gone. If your materials reference these, they are outdated.
• Treating every AI question as an AI question. Many questions that mention AI are actually testing cloud fundamentals like latency, throughput, or cost. The AI framing is just the scenario.
• Not knowing the Well-Architected Framework. It is used as evaluation criteria across the entire exam now. Knowing the six pillars and how they interact is essential.
• Ignoring Terraform/IaC. The v6.1 update made it explicit. You should be comfortable with basic Terraform resource provisioning.
• Memorizing case study details instead of practicing architectural reasoning. The exam tests how you think, not what you remember about a fictional company's org chart.
• Choosing self-managed solutions when a managed service works. The exam consistently rewards managed-service-first thinking.

What are the three biggest preparation mistakes?

1. Studying service features without studying architecture patterns. PCA tests judgment. Knowing that Cloud Spanner offers global consistency is table stakes. Knowing when it is a better choice than Cloud SQL with read replicas, and being able to explain why, is what the exam actually measures.
2. Not practicing case study reasoning under time pressure. Case studies consume more time per question than standalone questions. Get comfortable reading a question, finding the binding constraint, and narrowing down your options quickly.
3. Using study materials from before November 2025. The October 2025 update changed roughly 30% of the exam content. Older materials may be missing Vertex AI, AlloyDB, Migration Center, Model Armor, and the Well-Architected Framework requirements. Always check the publication date.

Is there a renewal option instead of retaking the full exam?

Yes. If you have an active PCA certification and you are within the renewal eligibility window (starting 60 days before expiration), you can take a shorter renewal exam. It is 1 hour, costs $100, has 25 questions, includes 1 case study focused on generative AI, and is almost entirely case study-based. If your certification has already expired, or if this is your first time getting certified, you will need to take the standard exam.

Where to Go After PCA

A few natural next steps depending on your area of interest:

• Google Cloud Professional Data Engineer for data architecture and ML pipeline roles.
• Google Cloud Professional Cloud DevOps Engineer for SRE and operations-focused roles.
• Google Cloud Professional Cloud Security Engineer for security architecture roles.
• Google Cloud Professional Cloud Network Engineer for networking-focused roles.
• Google Cloud Professional Machine Learning Engineer for AI/ML architecture roles.

Quick-Reference: What Is on the Exam (v6.1, October 30, 2025)

This is a condensed summary. For the full, always-current list, visit the official PCA exam guide.