How to Build an AWS Certification Study Plan

Step 4: Design the weekly loop — read, build, test

The engine of the plan is a loop every study week repeats, in three parts:

Read/watch (about half the week's hours)

Work through your chosen course or documentation for that week's domain. One primary resource per exam is enough; collecting three courses mostly produces the comfortable feeling of progress without the discomfort of recall. AWS's own Skill Builder platform is a sensible anchor here, since its exam-aligned content tracks the official guides.

Build (about a third)

Take the services highlighted in Step 1 and touch them for real: create the thing, break the thing, read the error, delete the thing. For an Associate-level candidate this might mean actually configuring the storage classes, load balancers or IAM policies that appeared in this week's material. Labs earn their place in the plan because exam scenarios describe operational situations — what happens when, what fails how — that reading alone never makes concrete. Mind your budget: build in a personal account with free-tier-eligible choices where possible, and tear resources down at the end of each session.

Test (the remainder, and never skipped)

Close each week with a short, untimed set of practice questions scoped to that week's domain, reviewing every answer — right ones included. This is a checkpoint, not a mock exam; its job is to catch the difference between "I watched it" and "I can apply it" while the material is a week old rather than a month old. The deeper craft of practice testing — when to start full mocks, what scores mean, how to review — is covered in the dedicated guide to using AWS practice tests effectively.

The loop's proportions flex with the exam. Cloud Practitioner candidates can shift hours from building towards reading, since the exam tests concepts more than configuration. Professional candidates should shift the other way, and add a fourth element: reading AWS architecture material critically, asking "why this design and not the alternative?", because that trade-off reasoning is precisely what SAP-C02 and DOP-C02 questions examine.

Step 5: Lay the weeks onto a calendar

With domains sized and the loop defined, the assembly is mechanical. A generic skeleton for a mid-sized window looks like this — stretch or compress the middle block to fit your own sizing from Step 3:

Phase	Weeks (adjustable)	Focus
Orientation	Week 1	Exam guide syllabus, self-audit, environment setup, resource choice
Domain cycles	Weeks 2–6	The read–build–test loop, one domain (or sub-domain) per cycle, heaviest weightings first
Consolidation	Week 7	Revisit every amber/red checkbox still unticked; re-do the labs that went badly
Benchmark and polish	Week 8	Full-length timed practice, weak-domain repair, logistics, rest

Three assembly rules make the difference between a calendar and a plan:

• Sequence by weighting, not by curiosity. Highest-weighted domains go earliest, when energy is freshest and there is still slack to overrun.
• Schedule sessions, not intentions. "Tuesday 19:00–20:30: IAM lab" survives a busy week; "study more this week" does not.
• Leave one deliberately empty slot per week. Overruns land there instead of cascading through the plan.

Step 6: Plan the final fortnight differently

The last two weeks are not more of the loop — they are a different mode. Content acquisition stops (nothing learned 48 hours before an exam is retrievable under pressure anyway) and evidence gathering starts:

1. Sit at least one full-length, timed practice exam under exam conditions — the full 90 minutes for a foundational exam, 130 for an Associate, up to 180 at Professional level. Timing behaviour is a skill of its own and this is where it gets rehearsed. A timed simulation, such as ExamPractice's practice-test mode, makes the conditions realistic rather than approximate.
2. Repair, don't re-cover. Feed the results back into your Step 1 checklist: wrong answers reopen specific checkboxes, and only those get restudied.
3. Handle logistics early. Choose between a Pearson VUE test centre and online proctoring, and if testing from home, run the system check on the actual machine and room you will use.
4. Understand the scoring before you walk in. AWS exams report a scaled score from 100 to 1,000 — the pass mark is 700 for foundational exams, 720 for Associate, 750 for Professional and Specialty — and a scaled score is not a percentage of questions correct. Knowing this stops both false confidence and false alarm when interpreting practice results.

Common study-plan mistakes (and the repair for each)

• Course-first planning. Symptom: your plan is a course's module list. Repair: rebuild it from the exam guide's domains; let the course serve the syllabus, not define it.
• The passive plateau. Symptom: weeks of watching, no building, no testing. Repair: enforce the loop — no week ends without a lab and a question set.
• Equal time for unequal domains. Symptom: five tidy weeks for five domains regardless of weighting or your audit colours. Repair: allocate by weighting × redness.
• Booking by bravado. Symptom: an exam date chosen to impress yourself, then rescheduled twice. Repair: size first (Step 3), book second.
• Studying for the wrong version. Symptom: 2023-era notes for a portfolio that has since changed — SysOps became CloudOps Engineer (SOA-C03) in late 2025, the Security Specialty moved to SCS-C03, and the Machine Learning Specialty retired in March 2026. Repair: confirm the current exam code on AWS's certification page the day you start planning, and again before you book.
• Skipping the consolidation weeks. Symptom: the plan ends on the last new domain. Repair: protect the final fortnight; it is where marginal candidates become passing ones.

A readiness checklist before you book

You are ready to schedule the exam when you can tick all of these:

• Every task statement from the exam guide is checked off or consciously triaged.
• You have hands-on familiarity with every in-scope service you flagged in Step 1 — at minimum, you have created and configured it once.
• A full-length timed practice run is complete, reviewed answer-by-answer, and your weakest domain has had a repair pass.
• You can explain, without notes, the core scenarios of the two highest-weighted domains to a colleague.
• Exam logistics (venue or proctoring setup, identification, fee and any discount voucher) are sorted.

If two or more boxes are empty, extend the plan by a week or two rather than gambling the fee — every AWS certification you do earn stays valid for three years, so a slightly later, safer pass costs you almost nothing in the long run.

Frequently asked questions

Should I build one plan for two exams at once?

Plan one exam at a time. Adjacent AWS exams share material, so your second plan's audit (Step 2) will start much greener — that is the efficient way to bank the overlap, rather than splitting focus across two syllabi.

Where do study groups or communities fit in the plan?

As an accountability layer, not a content source. A weekly check-in where you explain what you built is a strong forcing function; explaining a topic aloud is also one of the fastest ways to find your own gaps.

Does this method change for the newer AWS exams with ordering, matching and case-study questions?

The construction method is identical; only the final-fortnight rehearsal changes. Those question types currently appear on the AI-focused exams (they debuted on AIF-C01 and MLA-C01), so candidates for those should make sure their timed practice includes unfamiliar formats rather than assuming everything is multiple choice.

Which exam should this plan point at, if I haven't decided?

That is a different question from how to study, and it depends on your role and experience — the guide to choosing your first AWS certification handles it for newcomers, and the AWS certification roadmap covers sequencing beyond that.

From plan to pass

A good AWS study plan is unglamorous by design: a syllabus lifted from the official exam guide, an honest audit, a window sized from your real capacity, a weekly loop that always ends in questions, and a final fortnight spent gathering evidence instead of content. Build it once, put it somewhere visible, and then do the only genuinely hard part — follow it on the weeks when motivation dips. The candidates who pass are rarely the ones who studied the most hours; they are the ones whose hours were pointed at the right domains in the right order.