Where I StartedJanuary: 435, eleventh percentile. To be fully transparent, I had no idea how to interpret that score. I hadn't studied in four years — I didn't even know what the GMAT interface looked like. It was more of a touch and feel, a sense check to figure out where I actually stood.
One of the reasons I went with
e-GMAT was that I needed something self-paced — I work in consulting, so live classes were never going to happen. But more than that, I didn't want to spend mental capacity planning what to study next. The Personalized Study Plan (PSP) — which takes your starting score, your target, and the weeks you have, and turns it into an ordered sequence — did exactly that. I blindly followed it, to be honest. I only went back to the course material later when I was plugging topic gaps.
Quant — The Low-Hanging FruitQuant was the section that scared me the most going in, which is funny because it ended up being my highest. The thing is, I knew there was a lot of potential there. It wasn't really the concepts or the logic — it was the approach to the problem, the timing. I called it a low-hanging fruit because I felt like I just had to work a little bit towards it to fix that leak.
So while my plan had me focusing on Verbal first, I kept doing Quant on the side because I could see the payoff. The
e-GMAT course's timed sectional mocks converted what I knew into speed under pressure — my data wasn't showing a concept gap, it was showing I was solving it the slow way, and that's a different fix. I drilled it until the faster approach became automatic. Quant went from 73 to 87 on test day, and I actually beat my best mock by eight points on the real exam.
The Habit I Couldn't SeeVerbal was a different kind of problem. I kept convincing myself an answer was right, and once I'd talked myself into it, I genuinely could not see why it was wrong. Every time I read my thinking process, I had confirmation bias, and it created this rift — okay, but this is not the correct answer, why am I consistently thinking like that?
The fix started with error logs. My mentors asked me to type out exactly what was going through my brain while attempting each question, and why I chose my answer over the right one. Pretty candidly, no filter. They analysed it from a more experienced lens, a third person point of view, and caught the pattern I couldn't: I wasn't testing answers, I was judging them by how thorough they sounded. Once I switched to negation-testing — can I break this? — instead of "does this sound complete," my accuracy on hard assumption questions went from stuck near 40 percent to over 60 within two days. I finished CR at the 95th percentile; while practising, CRs were a lot harder for me than RCs, so that felt good.
Data Insights — Wordy When It CountedDI was the section I expected to be shakiest — 65, 8th percentile at my diagnostic — and it turned around the most. The structured DI course and its different question formats, reinforced by cementing quizzes — short, exam-level checks that force a skill to stick before you move on — plus DI sectional mocks built the timing I needed.
On test day, DI was extremely verbose. A lot of non-math questions, very wordy Multi-Source Reasoning passages with so much information, but they needed the most basic data to actually answer the question. It was more about effective test-taking than the actual content. I trained for exactly that — find the two numbers that matter, ignore the rest — and finished with two and a half minutes to spare. DI went from the 8th percentile to the 89th.
Having Someone Else Read My DataI'll be very honest: my data was confusing. After plugging topic gaps, I didn't know where I was losing marks. When my mentor Jagadish came on after a mock in the mid-600s, the biggest shift wasn't more content — it was someone outside my head reading the same error logs and immediately naming a pattern I'd been circling for weeks.
I was almost in two minds about whether to delay my exam. We had a conversation and I said I think I'm stuck and I'm just going to write it. That reassurance — someone confirming the plateau was real but telling me I was closer than it felt — mattered as much as the diagnosis. If you're wondering whether Last Mile Push mentorship is worth it, this is the case: not new material, a second pair of eyes on data you can't read objectively when it's your own.
The Mocks That Wouldn't BehaveMy practice scores did not climb in a straight line. I have a screenshot of two weeks of Quant attempts: 81, 82, 81, 82. The needle was just not moving and I felt very demotivated. Then I jumped from 545 to 615 in six days. And then — the mock before the actual exam was official mock 5, and I scored a 605. I broke down completely. Test conditions at home weren't great, there was commotion, and I was shell shocked.
I went into the real test telling myself it doesn't matter, I can take it again — that's the beauty of the GMAT. And then I scored a 695, above every mock that had rattled me. I was averaging 645 to 675 in mocks, so this was genuinely better than what I expected.
Making Peace With ItMy PSP target was 735, and 695 lands under it. But bottom-tenth to top-three-percent in under six months is not a consolation prize. The one thing I'd tell the version of myself staring at that 435: be consistent, trust the process even when the score isn't moving, and give yourself credit for what the climb actually cost.