AWS CLF-C02 Cloud Practitioner Readiness Matrix: How to Diagnose Your Weakest Exam Domains

 

AWS CLF-C02 Cloud Practitioner readiness is best approached as a decision-and-application problem rather than a collection of isolated facts. AWS weights the current CLF-C02 domains unequally: Cloud Concepts 24%, Security and Compliance 30%, Cloud Technology and Services 34%, and Billing, Pricing, and Support 12%. For AWS CLF-C02 Cloud Practitioner readiness, the preparation target is concrete: what you need to understand, how that understanding appears in scenarios, and what evidence shows that the knowledge is becoming reliable.

For AWS CLF-C02 Cloud Practitioner readiness, current official information matters because certification blueprints and product scope change. The official guide uses an overall compensatory score, so a domain matrix is a prioritization tool rather than a claim that each domain must be “passed” separately. For AWS CLF-C02 Cloud Practitioner readiness, those details provide boundaries rather than a shortcut, and they show how to allocate attention without studying the wrong material at the wrong depth.

The matrix below scores observable reasoning, then combines domain weight, error type, and transfer ability so you can decide where another hour of study is most valuable. Throughout this AWS CLF-C02 Cloud Practitioner readiness article, readiness is treated as evidence you can produce—an explanation, a correct comparison, a small practical exercise, or a defensible scenario decision. No checklist for AWS CLF-C02 Cloud Practitioner readiness can guarantee an exam result, but a good one can expose exactly what still needs work.

Build the Matrix Around Evidence, Not Confidence

For build the matrix around evidence, not confidence, this is where shallow recall usually breaks down and structured reasoning becomes more valuable. In a readiness matrix, build the matrix around evidence, not confidence should measure your ability to classify, explain, apply, and reject a close alternative in a fresh scenario. A build the matrix around evidence, not confidence score is only useful when it describes observable capability. For build the matrix around evidence, not confidence, replace vague labels such as ‘good’ or ‘bad’ with evidence: can you classify a scenario, explain competing choices, and predict the consequence of the selected service or concept without relying on a memorized keyword?

In a scenario, after reading about elasticity, ask yourself to distinguish it from scalability and high availability in three different business situations. For build the matrix around evidence, not confidence, rate the result on a simple evidence scale: 0 for unfamiliar, 1 for recognition after seeing options, 2 for unaided explanation, and 3 for application to a new scenario with a close distractor rejected. That evidence scale makes the build the matrix around evidence, not confidence row actionable rather than merely motivational.

The distinction to watch is recognition after seeing the answer is weaker evidence than producing the decision rule before viewing options. For build the matrix around evidence, not confidence, record errors by cause—knowledge gap, service confusion, responsibility-boundary error, pricing/support confusion, or rushed reading—because identical percentages can hide very different study needs.

For build the matrix around evidence, not confidence, add one diagnostic question to your study log: what would I need to observe before I could make this decision confidently? Repeat the build the matrix around evidence, not confidence diagnostic after a focused learning block and look for a change in reasoning quality, not just a higher number. Durable improvement in build the matrix around evidence, not confidence is visible when the rule transfers to a different scenario without reproducing the wording of the study material.

Score Cloud Concepts at the Decision Level

For score cloud concepts at the decision level, this topic rewards candidates who can move from definition to consequence without guessing. In a readiness matrix, score cloud concepts at the decision level should measure cloud value propositions, deployment economics, elasticity, scalability, high availability, and the AWS value framework as choices tied to business needs. A score cloud concepts at the decision level score is only useful when it describes observable capability. For score cloud concepts at the decision level, replace vague labels such as ‘good’ or ‘bad’ with evidence: can you classify a scenario, explain competing choices, and predict the consequence of the selected service or concept without relying on a memorized keyword?

When troubleshooting or choosing between alternatives, a startup with uncertain demand values a different cloud benefit than a regulated organization prioritizing resilience and governance. For score cloud concepts at the decision level, rate the result on a simple evidence scale: 0 for unfamiliar, 1 for recognition after seeing options, 2 for unaided explanation, and 3 for application to a new scenario with a close distractor rejected. That evidence scale makes the score cloud concepts at the decision level row actionable rather than merely motivational.

The distinction to watch is business benefit statements and implementation details live at different depths in a foundational exam. For score cloud concepts at the decision level, record errors by cause—knowledge gap, service confusion, responsibility-boundary error, pricing/support confusion, or rushed reading—because identical percentages can hide very different study needs.

For score cloud concepts at the decision level, turn this into a short worksheet: write the requirement, the likely component or concept, the evidence you would inspect, and the condition that would make your first choice wrong. Repeat the score cloud concepts at the decision level diagnostic after a focused learning block and look for a change in reasoning quality, not just a higher number. Durable improvement in score cloud concepts at the decision level is visible when the rule transfers to a different scenario without reproducing the wording of the study material.

Diagnose Security and Compliance Separately

For diagnose security and compliance separately, the practical point is not the label itself but the decision it changes. In a readiness matrix, diagnose security and compliance separately should measure shared responsibility, IAM concepts, data protection, governance, and compliance responsibilities rather than generic “security” familiarity. A diagnose security and compliance separately score is only useful when it describes observable capability. For diagnose security and compliance separately, replace vague labels such as ‘good’ or ‘bad’ with evidence: can you classify a scenario, explain competing choices, and predict the consequence of the selected service or concept without relying on a memorized keyword?

A useful contrast is that decide whether AWS, the customer, or both have responsibility for a control in a managed-service scenario and explain why. For diagnose security and compliance separately, rate the result on a simple evidence scale: 0 for unfamiliar, 1 for recognition after seeing options, 2 for unaided explanation, and 3 for application to a new scenario with a close distractor rejected. That evidence scale makes the diagnose security and compliance separately row actionable rather than merely motivational.

The distinction to watch is security service recognition is not the same as knowing who configures, operates, and governs the relevant control. For diagnose security and compliance separately, record errors by cause—knowledge gap, service confusion, responsibility-boundary error, pricing/support confusion, or rushed reading—because identical percentages can hide very different study needs.

For diagnose security and compliance separately, build a two-column contrast between the correct use case and the nearest plausible alternative; this is more valuable than adding another page of definitions. Repeat the diagnose security and compliance separately diagnostic after a focused learning block and look for a change in reasoning quality, not just a higher number. Durable improvement in diagnose security and compliance separately is visible when the rule transfers to a different scenario without reproducing the wording of the study material.

Weight Cloud Technology and Services Most Heavily

For weight cloud technology and services most heavily, the exam-oriented question is what evidence would make one option more appropriate than another. In a readiness matrix, weight cloud technology and services most heavily should measure service-category recognition and fit across compute, storage, databases, networking, analytics, AI/ML, management, and migration at the depth expected for a foundational role. A weight cloud technology and services most heavily score is only useful when it describes observable capability. For weight cloud technology and services most heavily, replace vague labels such as ‘good’ or ‘bad’ with evidence: can you classify a scenario, explain competing choices, and predict the consequence of the selected service or concept without relying on a memorized keyword?

When the wording changes, choose between object, block, and file storage from a requirement, or between common compute models based on operational responsibility. For weight cloud technology and services most heavily, rate the result on a simple evidence scale: 0 for unfamiliar, 1 for recognition after seeing options, 2 for unaided explanation, and 3 for application to a new scenario with a close distractor rejected. That evidence scale makes the weight cloud technology and services most heavily row actionable rather than merely motivational.

The distinction to watch is knowing a service name is weaker than mapping a requirement to the correct service category and explaining the trade-off. For weight cloud technology and services most heavily, record errors by cause—knowledge gap, service confusion, responsibility-boundary error, pricing/support confusion, or rushed reading—because identical percentages can hide very different study needs.

For weight cloud technology and services most heavily, create one lab or paper exercise that forces you to predict the result before checking documentation or a console, then record why your prediction was right or wrong. Repeat the weight cloud technology and services most heavily diagnostic after a focused learning block and look for a change in reasoning quality, not just a higher number. Durable improvement in weight cloud technology and services most heavily is visible when the rule transfers to a different scenario without reproducing the wording of the study material.

Treat Billing, Pricing, and Support as Precision Work

For treat billing, pricing, and support as precision work, the important distinction is between recognizing the term and being able to use it in context. In a readiness matrix, treat billing, pricing, and support as precision work should measure pricing models, cost-management tools, support options, and billing concepts where small wording differences can change the answer. A treat billing, pricing, and support as precision work score is only useful when it describes observable capability. For treat billing, pricing, and support as precision work, replace vague labels such as ‘good’ or ‘bad’ with evidence: can you classify a scenario, explain competing choices, and predict the consequence of the selected service or concept without relying on a memorized keyword?

Operationally, separate a tool that analyzes spend from a pricing commitment that changes the effective rate, and from a support plan that changes response access. For treat billing, pricing, and support as precision work, rate the result on a simple evidence scale: 0 for unfamiliar, 1 for recognition after seeing options, 2 for unaided explanation, and 3 for application to a new scenario with a close distractor rejected. That evidence scale makes the treat billing, pricing, and support as precision work row actionable rather than merely motivational.

The distinction to watch is cost visibility, cost optimization, pricing mechanism, and technical support are related but not interchangeable. For treat billing, pricing, and support as precision work, record errors by cause—knowledge gap, service confusion, responsibility-boundary error, pricing/support confusion, or rushed reading—because identical percentages can hide very different study needs.

For treat billing, pricing, and support as precision work, use a compact error log for this topic so that every missed question produces a rule you can apply to a different scenario rather than a fact you merely re-read. Repeat the treat billing, pricing, and support as precision work diagnostic after a focused learning block and look for a change in reasoning quality, not just a higher number. Durable improvement in treat billing, pricing, and support as precision work is visible when the rule transfers to a different scenario without reproducing the wording of the study material.

Add a Responsibility-Boundary Column

For add a responsibility-boundary column, instead of memorizing a sentence, connect the idea to what an administrator or decision-maker would actually observe. In a readiness matrix, add a responsibility-boundary column should measure whether you can place operational tasks correctly between AWS and the customer for IaaS-like and more managed services. A add a responsibility-boundary column score is only useful when it describes observable capability. For add a responsibility-boundary column, replace vague labels such as ‘good’ or ‘bad’ with evidence: can you classify a scenario, explain competing choices, and predict the consequence of the selected service or concept without relying on a memorized keyword?

Operationally, compare patching responsibility for an EC2 guest operating system with responsibility boundaries for a managed database service. For add a responsibility-boundary column, rate the result on a simple evidence scale: 0 for unfamiliar, 1 for recognition after seeing options, 2 for unaided explanation, and 3 for application to a new scenario with a close distractor rejected. That evidence scale makes the add a responsibility-boundary column row actionable rather than merely motivational.

The distinction to watch is managed service reduces some customer operations but does not make the customer free of data, identity, configuration, and governance duties. For add a responsibility-boundary column, record errors by cause—knowledge gap, service confusion, responsibility-boundary error, pricing/support confusion, or rushed reading—because identical percentages can hide very different study needs.

For add a responsibility-boundary column, rehearse it by explaining the idea without notes, then solve one scenario in which the obvious keyword is deliberately misleading. Repeat the add a responsibility-boundary column diagnostic after a focused learning block and look for a change in reasoning quality, not just a higher number. Durable improvement in add a responsibility-boundary column is visible when the rule transfers to a different scenario without reproducing the wording of the study material.

Add a Service-Selection Column

For add a service-selection column, a strong candidate treats this as a relationship between components rather than an isolated fact. In a readiness matrix, add a service-selection column should measure whether you can identify the requirement first and then choose the appropriate service family without keyword matching. A add a service-selection column score is only useful when it describes observable capability. For add a service-selection column, replace vague labels such as ‘good’ or ‘bad’ with evidence: can you classify a scenario, explain competing choices, and predict the consequence of the selected service or concept without relying on a memorized keyword?

In a scenario, a question mentioning “database” can still require a relational, key-value, warehouse, or other service category depending on the data and access pattern. For add a service-selection column, rate the result on a simple evidence scale: 0 for unfamiliar, 1 for recognition after seeing options, 2 for unaided explanation, and 3 for application to a new scenario with a close distractor rejected. That evidence scale makes the add a service-selection column row actionable rather than merely motivational.

The distinction to watch is category fit should be established before memorizing individual features of every AWS service. For add a service-selection column, record errors by cause—knowledge gap, service confusion, responsibility-boundary error, pricing/support confusion, or rushed reading—because identical percentages can hide very different study needs.

For add a service-selection column, turn this into a short worksheet: write the requirement, the likely component or concept, the evidence you would inspect, and the condition that would make your first choice wrong. Repeat the add a service-selection column diagnostic after a focused learning block and look for a change in reasoning quality, not just a higher number. Durable improvement in add a service-selection column is visible when the rule transfers to a different scenario without reproducing the wording of the study material.

Track Distractor Attraction

For track distractor attraction, a useful way to make the topic durable is to connect it to a concrete failure, constraint, or trade-off. In a readiness matrix, track distractor attraction should measure which wrong answers repeatedly feel plausible and what false rule is pulling you toward them. A track distractor attraction score is only useful when it describes observable capability. For track distractor attraction, replace vague labels such as ‘good’ or ‘bad’ with evidence: can you classify a scenario, explain competing choices, and predict the consequence of the selected service or concept without relying on a memorized keyword?

From a readiness perspective, if you repeatedly choose a scalable solution when the requirement is availability across failure domains, the problem is a concept boundary rather than a missing service name. For track distractor attraction, rate the result on a simple evidence scale: 0 for unfamiliar, 1 for recognition after seeing options, 2 for unaided explanation, and 3 for application to a new scenario with a close distractor rejected. That evidence scale makes the track distractor attraction row actionable rather than merely motivational.

The distinction to watch is wrong-answer patterns can reveal a stable misconception even when your overall score appears acceptable. For track distractor attraction, record errors by cause—knowledge gap, service confusion, responsibility-boundary error, pricing/support confusion, or rushed reading—because identical percentages can hide very different study needs.

For track distractor attraction, practice the topic in both directions: identify the right answer from a requirement, and infer the requirement from a proposed design or operational action. Repeat the track distractor attraction diagnostic after a focused learning block and look for a change in reasoning quality, not just a higher number. Durable improvement in track distractor attraction is visible when the rule transfers to a different scenario without reproducing the wording of the study material.

Separate Recall Errors From Reading Errors

For separate recall errors from reading errors, the highest-value study move is to turn the concept into a small decision model. In a readiness matrix, separate recall errors from reading errors should measure whether a miss occurred because knowledge was absent or because decisive wording was overlooked. A separate recall errors from reading errors score is only useful when it describes observable capability. For separate recall errors from reading errors, replace vague labels such as ‘good’ or ‘bad’ with evidence: can you classify a scenario, explain competing choices, and predict the consequence of the selected service or concept without relying on a memorized keyword?

At the boundary between topics, you may know that Reserved Instances and Savings Plans relate to discounted usage yet misread a question asking specifically about flexibility or commitment scope. For separate recall errors from reading errors, rate the result on a simple evidence scale: 0 for unfamiliar, 1 for recognition after seeing options, 2 for unaided explanation, and 3 for application to a new scenario with a close distractor rejected. That evidence scale makes the separate recall errors from reading errors row actionable rather than merely motivational.

The distinction to watch is content remediation and question-reading remediation require different practice methods. For separate recall errors from reading errors, record errors by cause—knowledge gap, service confusion, responsibility-boundary error, pricing/support confusion, or rushed reading—because identical percentages can hide very different study needs.

For separate recall errors from reading errors, explain the concept to an imaginary teammate using one normal case and one edge case; if the explanation depends on a memorized slogan, study the mechanism again. Repeat the separate recall errors from reading errors diagnostic after a focused learning block and look for a change in reasoning quality, not just a higher number. Durable improvement in separate recall errors from reading errors is visible when the rule transfers to a different scenario without reproducing the wording of the study material.

Measure Explanation Quality

For measure explanation quality, for preparation purposes, this area becomes useful when it is tied to an operational choice. In a readiness matrix, measure explanation quality should measure whether you can state why the chosen option fits and why the nearest alternative does not. A measure explanation quality score is only useful when it describes observable capability. For measure explanation quality, replace vague labels such as ‘good’ or ‘bad’ with evidence: can you classify a scenario, explain competing choices, and predict the consequence of the selected service or concept without relying on a memorized keyword?

In a scenario, after every practice item, explain the requirement in your own words before explaining the answer. For measure explanation quality, rate the result on a simple evidence scale: 0 for unfamiliar, 1 for recognition after seeing options, 2 for unaided explanation, and 3 for application to a new scenario with a close distractor rejected. That evidence scale makes the measure explanation quality row actionable rather than merely motivational.

The distinction to watch is correct guessing and correct reasoning produce the same percentage but very different readiness. For measure explanation quality, record errors by cause—knowledge gap, service confusion, responsibility-boundary error, pricing/support confusion, or rushed reading—because identical percentages can hide very different study needs.

For measure explanation quality, turn this into a short worksheet: write the requirement, the likely component or concept, the evidence you would inspect, and the condition that would make your first choice wrong. Repeat the measure explanation quality diagnostic after a focused learning block and look for a change in reasoning quality, not just a higher number. Durable improvement in measure explanation quality is visible when the rule transfers to a different scenario without reproducing the wording of the study material.

Use Weighted Priority, Not Raw Weakness

For use weighted priority, not raw weakness, this is where shallow recall usually breaks down and structured reasoning becomes more valuable. In a readiness matrix, use weighted priority, not raw weakness should measure combining domain weight, error frequency, and severity so limited time goes where it changes expected performance most. A use weighted priority, not raw weakness score is only useful when it describes observable capability. For use weighted priority, not raw weakness, replace vague labels such as ‘good’ or ‘bad’ with evidence: can you classify a scenario, explain competing choices, and predict the consequence of the selected service or concept without relying on a memorized keyword?

A useful contrast is that a recurring misconception in the 34% technology/services domain may deserve attention before a rare issue in the 12% billing domain. For use weighted priority, not raw weakness, rate the result on a simple evidence scale: 0 for unfamiliar, 1 for recognition after seeing options, 2 for unaided explanation, and 3 for application to a new scenario with a close distractor rejected. That evidence scale makes the use weighted priority, not raw weakness row actionable rather than merely motivational.

The distinction to watch is priority is not simply your lowest percentage; it is the expected value of fixing the weakness. For use weighted priority, not raw weakness, record errors by cause—knowledge gap, service confusion, responsibility-boundary error, pricing/support confusion, or rushed reading—because identical percentages can hide very different study needs.

For use weighted priority, not raw weakness, add one diagnostic question to your study log: what would I need to observe before I could make this decision confidently? Repeat the use weighted priority, not raw weakness diagnostic after a focused learning block and look for a change in reasoning quality, not just a higher number. Durable improvement in use weighted priority, not raw weakness is visible when the rule transfers to a different scenario without reproducing the wording of the study material.

Retest With New Wording

For retest with new wording, this topic rewards candidates who can move from definition to consequence without guessing. In a readiness matrix, retest with new wording should measure whether improved performance transfers to scenarios that do not resemble the material used for correction. A retest with new wording score is only useful when it describes observable capability. For retest with new wording, replace vague labels such as ‘good’ or ‘bad’ with evidence: can you classify a scenario, explain competing choices, and predict the consequence of the selected service or concept without relying on a memorized keyword?

The risk of treating this superficially is that after studying shared responsibility, answer scenarios involving compute, storage, managed databases, and account security rather than repeating the same example. For retest with new wording, rate the result on a simple evidence scale: 0 for unfamiliar, 1 for recognition after seeing options, 2 for unaided explanation, and 3 for application to a new scenario with a close distractor rejected. That evidence scale makes the retest with new wording row actionable rather than merely motivational.

The distinction to watch is memorized wording creates false improvement; transfer demonstrates a more stable concept. For retest with new wording, record errors by cause—knowledge gap, service confusion, responsibility-boundary error, pricing/support confusion, or rushed reading—because identical percentages can hide very different study needs.

For retest with new wording, use a compact error log for this topic so that every missed question produces a rule you can apply to a different scenario rather than a fact you merely re-read. Repeat the retest with new wording diagnostic after a focused learning block and look for a change in reasoning quality, not just a higher number. Durable improvement in retest with new wording is visible when the rule transfers to a different scenario without reproducing the wording of the study material.

Add Time and Cognitive Load

For add time and cognitive load, the practical point is not the label itself but the decision it changes. In a readiness matrix, add time and cognitive load should measure whether your reasoning remains dependable when several domains are mixed and you are working at realistic pace. A add time and cognitive load score is only useful when it describes observable capability. For add time and cognitive load, replace vague labels such as ‘good’ or ‘bad’ with evidence: can you classify a scenario, explain competing choices, and predict the consequence of the selected service or concept without relying on a memorized keyword?

During review, use small timed sets that combine security, service selection, and pricing instead of isolating every topic forever. For add time and cognitive load, rate the result on a simple evidence scale: 0 for unfamiliar, 1 for recognition after seeing options, 2 for unaided explanation, and 3 for application to a new scenario with a close distractor rejected. That evidence scale makes the add time and cognitive load row actionable rather than merely motivational.

The distinction to watch is speed without accuracy is unhelpful, but accuracy that collapses under normal time pressure still needs rehearsal. For add time and cognitive load, record errors by cause—knowledge gap, service confusion, responsibility-boundary error, pricing/support confusion, or rushed reading—because identical percentages can hide very different study needs.

For add time and cognitive load, build a two-column contrast between the correct use case and the nearest plausible alternative; this is more valuable than adding another page of definitions. Repeat the add time and cognitive load diagnostic after a focused learning block and look for a change in reasoning quality, not just a higher number. Durable improvement in add time and cognitive load is visible when the rule transfers to a different scenario without reproducing the wording of the study material.

Create a Red-Amber-Green Exit Rule

For create a red-amber-green exit rule, the exam-oriented question is what evidence would make one option more appropriate than another. In a readiness matrix, create a red-amber-green exit rule should measure clear thresholds for what counts as unknown, fragile, and reliable so your study plan can change based on evidence. A create a red-amber-green exit rule score is only useful when it describes observable capability. For create a red-amber-green exit rule, replace vague labels such as ‘good’ or ‘bad’ with evidence: can you classify a scenario, explain competing choices, and predict the consequence of the selected service or concept without relying on a memorized keyword?

During review, a green topic should survive an explanation test and a new scenario; amber needs targeted review; red needs concept rebuilding. For create a red-amber-green exit rule, rate the result on a simple evidence scale: 0 for unfamiliar, 1 for recognition after seeing options, 2 for unaided explanation, and 3 for application to a new scenario with a close distractor rejected. That evidence scale makes the create a red-amber-green exit rule row actionable rather than merely motivational.

The distinction to watch is color labels are only useful when each color is tied to a repeatable test. For create a red-amber-green exit rule, record errors by cause—knowledge gap, service confusion, responsibility-boundary error, pricing/support confusion, or rushed reading—because identical percentages can hide very different study needs.

For create a red-amber-green exit rule, practice the topic in both directions: identify the right answer from a requirement, and infer the requirement from a proposed design or operational action. Repeat the create a red-amber-green exit rule diagnostic after a focused learning block and look for a change in reasoning quality, not just a higher number. Durable improvement in create a red-amber-green exit rule is visible when the rule transfers to a different scenario without reproducing the wording of the study material.

Turn the Matrix Into a Final Study Queue

For turn the matrix into a final study queue, the important distinction is between recognizing the term and being able to use it in context. In a readiness matrix, turn the matrix into a final study queue should measure a ranked list of specific misconceptions and scenario types rather than a broad instruction to “study AWS”. A turn the matrix into a final study queue score is only useful when it describes observable capability. For turn the matrix into a final study queue, replace vague labels such as ‘good’ or ‘bad’ with evidence: can you classify a scenario, explain competing choices, and predict the consequence of the selected service or concept without relying on a memorized keyword?

From a readiness perspective, schedule the top three weaknesses with a concrete resource or exercise and a retest date, then stop spending equal time everywhere. For turn the matrix into a final study queue, rate the result on a simple evidence scale: 0 for unfamiliar, 1 for recognition after seeing options, 2 for unaided explanation, and 3 for application to a new scenario with a close distractor rejected. That evidence scale makes the turn the matrix into a final study queue row actionable rather than merely motivational.

The distinction to watch is a diagnostic becomes valuable only when it changes the next action. For turn the matrix into a final study queue, record errors by cause—knowledge gap, service confusion, responsibility-boundary error, pricing/support confusion, or rushed reading—because identical percentages can hide very different study needs.

For turn the matrix into a final study queue, add one diagnostic question to your study log: what would I need to observe before I could make this decision confidently? Repeat the turn the matrix into a final study queue diagnostic after a focused learning block and look for a change in reasoning quality, not just a higher number. Durable improvement in turn the matrix into a final study queue is visible when the rule transfers to a different scenario without reproducing the wording of the study material.

Putting the Preparation Into Practice

At the end of , rank weighted domain evidence, error causes, transfer, and scenario performance by expected payoff from one more hour of study. Keep the ranking tied to fresh evidence so a comfortable topic does not keep taking time from a weaker but more important domain.

Recalculate the matrix after each major review cycle and retire topics from the queue only when new wording no longer causes the same mistake. The final check for is transfer: a new scenario should still produce the right rule and a clear reason the closest distractor fails.

Use the AWS CLF-C02 practice-test page as one source of diagnostic evidence, not as a score-chasing exercise. Tag each miss to the matrix category that caused it and use the explanation to design the next targeted review.