GMAT 715 (V86, Q90, DI82) — CR from the 37th Percentile to a V86 — with e-GMATI was already grinding the hardest questions I could find. What I did not have was any idea what to change.715 (V86, Q90, DI82), 99th percentile
Where I actually wasMy internship ended on 17 July and the plan was four weeks of prep, then the exam. Every senior said the same thing. You cleared JEE, quant is easy, DI is easy, only verbal needs work. So I booked for 17 August, and on 30 July a practice mock came back at 695. It did not feel like good news. The paper felt easy, seniors confirmed early mocks run inflated, and for Harvard, Stanford and Wharton with a scholarship I needed 705 plus. I was already doing 705 to 805 level questions on GMAT Club and could not have told you which twenty points I was chasing. One word for that state: clueless.
Why I went looking for someone to read my dataI did not need more questions. I needed someone to do the strategising so my hours went into executing instead of guessing. That is why I registered for the Last Mile Push,
e-GMAT's closing-phase mentorship where a mentor works through your actual data and finds the behavioural patterns capping your score, and I moved my exam to 29 August.
Verbal, which was the real gapOn medium verbal, given enough time, I got it right, so medium was speed and speed came from volume. Hard was different, and I was convinced I was missing concepts. I was not.
After my first two hard cementing quizzes in the
e-GMAT course, which are short exam-level quizzes at the end of a topic with a threshold you clear or re-clear, I stopped chasing concepts and broke every question down through the process instead. What it asked, what the conclusion was, my pre-thinking, then the answer choice analysis. Pre-thinking means deciding what the right answer has to do before you look at the options. Writing that out question by question was slow, and it was the turn. My errors were not spread across topics; they collapsed into two repeating buckets, and I drilled those two until the check fired on its own. CR went from the 37th percentile to a V86, the 96th percentile.
Boldface was the one place I truly lacked concepts, so I did the relevant modules, and what they fixed was vocabulary. Assumption, prediction, claim, evidence, position taken by the author. Once those mean something precise, you kill an option the moment it calls something evidence that is not evidence. RC was already a strength but my notes were a mess, and the modules gave me a structural map, two lines a paragraph, what it is about and what it is doing.
Quant, where I assumed there was nothing to fixI was vouching for a Q90 and barely scoring full marks anywhere; sectionals sat at 82 and 87, with 88 my ceiling. The sectional mocks in the
e-GMAT course were a mirror. They are built from medium and hard questions only, so the real exam would be gentler than my practice, and I had more time per question than I believed.
Then I bucketed my mistakes and found exactly three. Missing a word or phrase in the question, my famous one. An at least, an at most, a greater than or equal to. Silly arithmetic; I have written 2 into 3 equals 5. And my handwriting, because my 5 comes back to me as a 6. Three fixes. Capture the constraint in writing before solving, write every step out, write numbers I can read.
Number properties was the real content gap, odd and even and primes, the block that got me during JEE too. I worked it accuracy first and speed second, since I could spend extra seconds there and claw them back where I was strong. By the end of that set I was at 100 percent accuracy on the block and the speed followed. 88 in sectionals became a Q90 on exam day.
Data Insights, kept for lastTwo-part analysis is CR wearing a different skin and multi-source reasoning is RC with tabs, so sequencing DI last meant everything carried over. TPA still caught me. Early cementing quizzes were good, then two went below the gate, and when I took another to prove I was fine I scored lower still. The gap was conceptual, so I went back into the DI modules and their application files and practised until a method existed, which was to read the answer choices once and let the mapping between the parts resolve before I solve. Two quizzes below the gate ended as a DI82, the 88th percentile.
Sectionals, then a 595 six days outMy first verbal sectional was 16 of 25 and a V81, with five questions left in the last five minutes and five of my seven considered errors avoidable by reading properly. I started a corrective action note and revised it daily, because you forget your own errors inside a week if they are not on paper.
My first full mock then came in at 595. I expected a wall of crosses; instead my accuracy and sub-topic percentiles were good, and what had actually happened was three unattempted questions in verbal and two in DI. The penalty for not completing a section is far heavier than the penalty for being wrong. That let me stay calm, because nothing was wrong with the process. Over the next four days I never left anything unanswered, ran custom quizzes in Scholaranium on efficient reading, put an upper time limit on every question with a flag and a review pass, and revised the corrective actions daily. The next mock was a 705.
What the mentorship actually didMy mentor Kavya could tell me which module to open when I had no time to go everywhere, but the part I could not have replicated alone was the read on my own behaviour. I once worked on verbal for four days, took the verbal cementing quiz to validate it, and scored badly enough that I wanted to throw the plan out. What came back was not comfort. It was that I had taken the quiz as the last thing in my day, after tiring my head out. I took it again the next morning with a clear head and got an 80. I never once scored a Q90 in a mock either, and saying that out loud instead of hiding it is what let me walk in still believing it was available. If you are short on time, look into LMP.
Exam dayFinal mock on Thursday, 705, which was the validation I wanted, and Friday I deliberately did nothing useful. Verbal was hard and DI was hard, which is usually good news. Going into the last section I told myself the only thing left was quant, quant is mine, everything else is done. 715.
Temporary dips are not signals; a bad quiz can just be a tired evening. And breaking one question down for half an hour taught me more about myself than a hundred I got right.
>> !!!
You do not have the required permissions to view the files attached to this post.