Why We Built SM-2 Into AzurePrep (And Why Most Cert Platforms Don't Bother)

Most exam prep platforms work like a slot machine. Answer 50 questions, get a score, come back tomorrow and see a random shuffle of the same pool. It feels like studying. It is not. The questions you got wrong last session are just as likely to appear on attempt 50 as they were on attempt 1. There is no mechanism that actually tracks what you know - only a counter that tracks how many times you clicked next.

AzurePrep is built differently. Every question you answer, right or wrong, feeds into a scheduling engine that decides exactly when you should see that concept again. This article explains how it works and why it matters for passing Azure certifications.

The Problem With Random Practice

The failure mode of random question banks is well-documented in learning research. You get comfortable with the questions you happen to see often. You forget the ones you saw three weeks ago. And on exam day, the distribution of topics you face bears no relationship to the distribution of topics you actually practised.

For Azure certifications specifically, this is a serious problem. AZ-104 alone covers identity, networking, storage, compute, and monitoring - each with dozens of subtopics. Random shuffling gives you false confidence in areas you happened to review recently and leaves gaps in areas you have not touched since week one.

The fix is not to study harder. It is to study at the right time.

What SM-2 Is

SM-2 is a spaced repetition algorithm developed in 1987 and still used in Anki, Duolingo, and dozens of other learning tools. It schedules each card individually based on how well you know it. Rate a card "easy" and you will not see it again for weeks. Rate it "again" and it comes back tomorrow. The interval grows each time you answer correctly, and shrinks when you struggle. Over time, review sessions stay short while long-term retention climbs.

How AzurePrep Uses It

There are two places SM-2 runs in AzurePrep:

Flashcard mode is the obvious one. Study a flashcard, rate your confidence, and SM-2 calculates your next review date. Cards you know well drift out to months-long intervals. Cards you keep missing stay close.

The Review Deck is the differentiator. When you take an exam simulation and get a question wrong, that question is automatically added to your SM-2 review queue. You do not have to do anything. The platform sees the wrong answer, creates a review card, and schedules your first review for the next day.

This means every exam attempt is also a diagnostic. You are not just checking your score - you are populating a personalised revision list that will surface your exact weak spots at the exact right time.

This is the part most platforms skip entirely. They show you a review screen at the end of an exam, but that review is passive - you read the explanation and move on. AzurePrep turns every wrong answer into an active scheduling event. The question goes into a queue, and the algorithm decides when you are ready to see it again.

What This Looks Like In Practice

The forgetting curve below shows why timing matters. Without review, you lose most of what you learned within a week. With SM-2 review sessions placed at the right intervals, retention stays high with far less time per session.

Each review session resets the decay clock. The key insight is that you do not need to review everything every day - only the cards that are due. That is what the scheduling engine handles.

Here is what interval growth actually looks like for a card you consistently rate "Easy":

A concept you genuinely understand will drift to intervals measured in months. You spend your time on what you do not know, not on what you have already mastered. After 5 reviews with an ease factor at or above 2.5 and three consecutive correct answers, a card is marked as mastered - it exits the active rotation unless you start getting it wrong again.

The Four Ratings

After each review card, you choose one of four ratings. The choice directly affects both the next interval and your ease factor - the multiplier that determines how fast intervals grow.

Ease factor starts at 2.5, has a floor of 1.3, and grows without a ceiling. If you consistently rate cards "Easy," the ease factor climbs and intervals compound faster. If you struggle and rate "Again" repeatedly, the ease factor drops toward 1.3 and intervals stay short - which is exactly what should happen when you do not know something well.

How to Get the Most Out of It

• 1 Do your reviews before starting new exam sessions. Each session surfaces up to 10 new cards plus all cards that are due. If you skip reviews and go straight to practice exams, the wrong answers pile up in your queue and the scheduling falls out of sync.
• 2 Rate honestly. The algorithm only works if your ratings reflect reality. Pressing "Easy" on something you guessed correctly will push it to a 35-day interval, and you will fail it on the real exam. Rate "Hard" or "Again" whenever there was any doubt.
• 3 Run exam simulations regularly, not just at the end of your prep. Every wrong answer is a signal that goes straight into your review queue. Running exams early in your study period means the scheduler has more time to surface and reinforce your weak spots before exam day.
• 4 Check your mastered card count, not just your practice exam score. A high score built on recently reviewed questions can be misleading. Mastered cards - those with 5 or more reviews, ease above 2.5, and a 3-review streak - are a more reliable signal of what will stick on exam day.

The Bottom Line

Spaced repetition is not a new idea. The research behind it has been solid for decades. What is different in AzurePrep is the integration: wrong exam answers do not vanish into a review screen you close and forget. They become scheduled events in a queue that follows the same algorithm as your deliberate flashcard study.

The result is a study system where every session - exam or flashcard - makes the next session more targeted. You spend less time re-reviewing things you already know, and more time closing the gaps that would cost you points on the real exam.

That is the difference between a question bank and a study engine.