I come from a tier-3 engineering college and currently work as an AI Engineer at National Payments Corporation of India with ~3 years of experience. Being a male engineer, I knew from the start that I would need a relatively higher GMAT score (around 675+) to stay competitive for ISB.
My first mock, taken without any preparation, was a
505 (Q80, V78, DI 67). At that time, I made the common mistake of thinking this was a decent starting point and that I could improve quickly. It didn’t take long to realize that improving 150+ points in GMAT is not a linear or easy process. It requires not just effort, but direction.
I began my preparation in June with the @GMAT Ninja 14-week plan. In hindsight, this was one of the best decisions I made early on. The biggest advantage of this plan is that it gives structure — something most beginners lack. Instead of randomly solving questions, I had a weekly roadmap, clear concepts to focus on, and a gradual build-up in difficulty. It also helped me understand how GMAT questions are designed rather than just what they are asking.
After about 7 weeks of preparation, I took another mock and scored
555 (Q83, V80, DI 70). This was a difficult moment. I had been putting in consistent effort and was doing reasonably well on practice benchmarks, but the score didn’t reflect that. Looking back, this phase was necessary because it forced me to confront a key gap: I was preparing for the GMAT, but I hadn’t yet learned how to take the GMAT.
Since I had limited time and wanted to apply in the same cycle for ISB, I decided to take mentorship through
GMATWhiz. This marked a shift in my preparation. Through sessions with Piyush, I realized that my concepts were largely in place, but my execution was inconsistent. I was not managing time well, I was second-guessing decisions, and most importantly, I didn’t have a clear test-taking strategy.
From that point on, my preparation became much more deliberate. I was studying around 15 hours per week, usually focusing on one section per day. Before my first attempt, I gave five mocks. The scores were fluctuating — I scored as high as 675 and as low as 585. Initially, this was frustrating, but instead of attributing it to randomness, I started analyzing the reasons behind each fluctuation.
For example, on the day I scored 585, I had poor sleep and there was background noise during the test. In another mock, I experimented with a different approach mid-exam, which disrupted my flow. Over time, I realized that these fluctuations were not due to lack of ability, but due to inconsistencies in execution. I made a list of such factors and ensured I eliminated them one by one. This gave me confidence that my true level was closer to my higher mock scores.
One important optimization I made was deciding the section order. After multiple trials, I settled on
Verbal → Quant → DI. Verbal and DI required more cognitive effort for me, while Quant felt relatively lighter. Placing Quant in the middle acted as a mental reset and helped me stay composed for the final section.
Quant was actually my strongest section in terms of concepts, but I was losing points due to careless mistakes. This is something both Charles (from GMAT Ninja) and Piyush emphasized repeatedly — the Quant section penalizes mistakes heavily. I changed my approach in two ways. First, I made it a rule to read every question twice before solving. Second, if a question was taking more than three minutes, I would leave it without hesitation. The focus shifted from attempting more questions to maintaining long streaks of correct answers. This change had a visible impact, and my Quant scores stabilized in the Q86–Q89 range in mocks.
Data Insights was the section where I saw the most improvement. Initially, I approached it with the mindset of completing all 20 questions. This backfired because in the pressure to finish, I made more mistakes. With guidance, I adopted a much simpler strategy: for every 15 minutes, aim to solve 5 questions correctly. This ensured that I stayed on track without rushing. It also reduced anxiety significantly. Over time, my DI scores moved from the high 60s/low 70s to consistently above 80.
Verbal was the section where I maintained consistency but did not see dramatic improvement. I stayed in the V80–82 range. Instead of over-optimizing here, I focused on avoiding bad performances and maintaining a stable baseline.
In my first official attempt in November, I scored
655 (Q86, V81, DI 80). While this was a solid improvement from where I started, I felt I had room to push a bit more, especially in Quant and DI.
Before my second attempt, I took two more mocks and scored
695 in both (Q88–Q89 range in Quant, DI in low-mid 80s). Going into the exam, I felt confident. However, the actual test didn’t go exactly as planned. I ended up scoring
665 (Q84, V81, DI 84). My Quant score dropped compared to mocks, even though it was my strongest section.
This was an important lesson. It reinforced that on any given day, one section can underperform. You cannot rely on a single strength — you need all sections to be stable enough to absorb variance. That realization changed how I viewed the exam entirely.
In the final phase of my preparation, I focused heavily on mocks and analysis. I gave mocks weekly, and closer to the exam, I increased the frequency (5 days before, 3 days before). I also used GMAT Club mocks to test specific strategies, especially around timing discipline and decision-making (like strictly not exceeding 3 minutes per question).
Looking back, a few things clearly worked for me. Having a structured plan in the beginning prevented me from wasting time. Mentorship helped me transition from learning to execution. Most importantly, analyzing mocks deeply — not just scores, but behavior — was the biggest differentiator.
At the same time, I made mistakes that are worth highlighting. Early in my preparation, I equated concept clarity with readiness, which is not true for GMAT. I also ignored Data Insights initially, thinking it would improve automatically. Another mistake was trying to attempt all questions, which reduced accuracy. It took time to understand that the GMAT rewards discipline more than aggression.
For anyone preparing, my biggest advice would be this: treat GMAT as a test of decision-making under constraints. And most importantly, don’t try to figure out your strategy during the exam — it should be fully defined before you walk in.
I decided to apply with
665 instead of taking another attempt because I felt the incremental improvement would require disproportionate effort, and I wanted to focus on applications within the available timelines.