Every CAT aspirant takes mocks. Very few analyse them properly - and that gap is the single biggest reason two students with the same number of mocks end up with wildly different percentiles.

Most students who feel like they've "analysed" a mock have really just looked at their score, skimmed the questions they got wrong, and moved on within 20-30 minutes - which isn't analysis, it's an emotional check-in, not a system for improvement (Quantifiers). This guide breaks down exactly what real mock analysis looks like, section by section, so you can build a repeatable system instead of guessing.

Step 1: Don't analyse immediately

Right after a mock, you're mentally and physically drained - and that's the worst state to analyse in. The right first move is a 30-60 minute break before you open a single solution (InsideIIM).

Once you're ready, do a fast first-pass triage: start with questions you got wrong that you thought you'd gotten right. These are your cheapest wins, because they're rarely conceptual gaps - usually a misread unit, a rushed calculation, or a skipped line, and fixing them takes minimal effort for real score gains (InsideIIM).

Step 2: Categorise every mistake - don't just note "wrong"

The single highest-leverage habit in CAT prep is tagging every wrong answer into a category, not just marking it incorrect. The four buckets that show up consistently across expert frameworks are:

  • Conceptual gap - the underlying topic genuinely needs revision
  • Careless mistake - a calculation or reading slip under pressure
  • Misread question - you understood the question incorrectly
  • Time pressure - you rushed into a wrong answer or guessed

Logging every wrong answer into one of these buckets after each mock is what builds pattern awareness over time (Cracku).

A more advanced version of this, used by test-prep mentors, adds two more categories worth tracking separately: the Time-Waster (a question that took disproportionately long relative to its difficulty) and the Black Hole (a question that took a long time and was still wrong) - both of which quietly damage your percentile more than an outright skip would (Career Launcher).

Step 3: Analyse each section differently - one method doesn't fit all three

VARC: Identify each passage's tone and category (science, economics, philosophy, narrative). A pattern across categories you consistently struggle with is far more useful than any single mock's score, and for every wrong answer you should be able to articulate the exact logical gap between your choice and the correct one - not just "I misread it" (Quantifiers). Most VARC errors specifically come down to overthinking or underreading (Career Launcher).

DILR: The real skill being tested isn't solving - it's set selection. After every mock, list every set in the order you attempted it, then rank the sets by their actual difficulty now that you've seen the full paper (Quantifiers). Spending 15 minutes on a set you never solve is often the single costliest decision in a CAT attempt (Career Launcher).

QA: Even questions you got right deserve a second look - check if there was a faster method or a shortcut using the answer options. If the same topic (say, Algebra) shows up wrong across three separate mocks, that's not a bad day - it's a structural weakness that needs you to pause mock-taking, revisit fundamentals, and only then return to testing (Career Launcher).

Step 4: Budget real time for analysis - not leftovers

Here's the number most aspirants get wrong: at least 40% of your total mock-test time should go into analysis, not just attempting. If a mock takes 2 hours, that means roughly 1.5-2 hours of review - and students who skip this step are the ones who typically plateau by August, while those who analyse consistently keep climbing right up to the exam (Stalwart Careers).

That's a significant chunk of time - which is exactly why analysis is the step most aspirants quietly cut when they're tired or short on time, and it's also exactly the step that determines whether their next ten mocks look any different from their last ten.

Step 5: Keep one running error log - not scattered notes

All three sections' analysis should feed into a single running sheet, not separate notes across different mocks - with each entry recording the topic, the mistake type, and the specific learning from it (Quantifiers, IIMking). Over weeks, this log becomes your personalised playbook: the same three or four errors keep resurfacing until you actually deal with them at the root.

This single log is also how you catch a plateau early. If a topic keeps appearing in your error log for three consecutive mocks, that's a structural gap hiding behind a flat score - and it's usually invisible if you're only looking at your overall percentile (Quantifiers).

A single mock tells you almost nothing on its own. What matters is the trend: is your rate of conceptual errors going down over time? Is your accuracy climbing while your attempts stay disciplined? Comparing your percentile and sectional scores across mocks - not fixating on any one result - is what actually tells you whether your prep is working (Stalwart Careers).

Research into stagnating scores backs this up directly: the primary drivers of a plateau are shallow, one-time analysis, the same error types recurring unnoticed across mocks, and concept gaps that never get formally diagnosed - not a lack of hours studied (Optima Learn).

Why most self-study aspirants stop doing this by mock #10

Everything above works - the problem is doing it by hand, consistently, for 25-30 mocks over several months. A manual error log demands real discipline: tagging every wrong answer, re-checking every "correct" one for a faster method, and cross-referencing three months of notes to spot a repeating pattern. Most aspirants do it diligently for the first few mocks and quietly stop by the time it matters most - right when patterns should be getting clearer, not harder to track (StudyRiser).

This is exactly the gap Marg was built to close. Instead of you manually tagging every wrong answer and cross-referencing a spreadsheet after each mock, Marg does the pattern-tracking for you - spotting which specific mistake type keeps repeating across your mocks and building your next day's VARC, DILR and QA practice directly around it. It's the "single running error log" this entire framework depends on, minus the manual upkeep.

If you've been taking mocks consistently but can't tell whether your last five actually taught you anything, that's worth checking. You can try Marg for free and see your own mistake pattern laid out in minutes.