How to improve from 60 to 95 percentile in CAT within the next 4 months?

Question

B.K Sharma · Accepted Answer

Went from 62 to 94.6 last year. Four months is enough, but only if you stop trying to &ldquo;cover everything.&rdquo; At 60 percentile, the issue is usually not concept gaps,  it&rsquo;s question selection, time management, and silly mistakes.
First, start writing weekly mocks. Not just for practice ,  the real game is in analysing them. Track what type of questions you keep getting wrong and stop touching them in actual mocks. That alone bumped my score up.
Second, fix your accuracy in VARC. Stop reading random RCs. Use past CAT RCs and AIMCATs only. Practice two a day, untimed first, then timed. For VA, only stick to para summary and jumbles if they&rsquo;re your strength.
In DILR, stopped trying to solve every set. Picked 3&ndash;4 familiar types and got better at spotting them in mocks. Choosing the right set made a bigger difference than finishing it fast. Quant was all about improving accuracy in easy and moderate questions. Didn&rsquo;t bother with advanced stuff , just aimed to get 18 to 20 questions right.

D. Aarav · Answer

Jumped from 61 to 96 last year, and the biggest shift came from changing how the prep was approached ,  not just what was studied.
&bull; Started by taking two mocks early on to find out where marks were being lost. VARC accuracy was below 50, QA timing was off, and wrong DILR sets were being attempted. &bull; Made VARC a daily habit. Practiced three RCs a day , one untimed to build flow, two timed to push speed. One day every week was just for VA ,  summaries, para jumbles, and odd-one-out. &bull; For QA, listed 25 important topics and stuck to those. Spent 90 minutes daily solving moderate-level questions and revised formulas weekly. No time wasted on rarely asked areas. &bull; DILR prep focused on familiar set types ,  arrangements, games and tournaments, bar-graph sets, and selection puzzles. One set per day from CAT 2017 to 2023 was enough. &bull; Took a full mock every five days. Two full days were kept for deep analysis , finding question types that consistently went wrong and figuring out why.
It wasn&rsquo;t about covering everything. It was about fixing weak patterns, building speed, and making every hour count.