e-GMAT: Stuck at 655 for Three Attempts — 675 with a Perfect Q90 after Last Mile Programg

Strengths:

Highly structured preparation: Clear module-based learning path—learn the concept, practice, identify weaknesses, and diagnose errors rather than blindly grinding questions.

Strong diagnostic approach: The platform helps distinguish between conceptual gaps, process errors, execution errors, and timing issues, making preparation targeted and data-driven.

Very realistic mocks: The five full-length mocks closely mirrored the actual GMAT in terms of difficulty and exam experience.

Challenging Quant question bank: Quant questions are intentionally harder than the actual exam, helping build confidence and making real-test questions feel more manageable.

Effective Verbal framework: Repeatable strategies, particularly pre-thinking in Critical Reasoning, helped create a consistent approach across question types.

Extensive sectional mocks: A large number of sectional tests allow students to practice individual sections and build stamina without taking a full-length mock every time.

Strong error-log integration: The course works particularly well with an error log, allowing students to identify recurring patterns and understand why scores are plateauing.

Data-driven preparation: Accuracy and performance data provide an objective basis for deciding what to study and what not to spend time on.

Excellent for score plateaus: Particularly useful for candidates who have already covered the concepts but are stuck at the same score despite repeated attempts.

Last Mile Push: Provides focused guidance during the final stage, when the problem is often exam strategy, consistency, stamina, and execution rather than knowledge.

Personalized mentorship: Mentor intervention can identify behavioral and preparation-related issues that are difficult to recognize through self-study. Kudos to my mentor, Subham.

Exam simulation: The combination of realistic mocks, sectional practice, and stamina-building exercises helps replicate the pressure and flow of the actual GMAT.

Holistic preparation: Compared with the alternatives you tried, e-GMAT combined content, practice, analytics, mocks, strategy, and mentorship in one ecosystem.

Proven through repeated attempts: Your experience of moving from a 575 diagnostic to 675 after multiple attempts, including breaking a three-test 655 plateau, demonstrates how the platform can support incremental score improvement.

Would make the product better:

Already more than sufficient.

I scored a 675 (Q90, V81, DI80) on the GMAT Focus Edition — a perfect quant score — after a journey that tested me more than most: five attempts across a year and a half, starting from a 575 diagnostic, with three consecutive scores of 655. I say this up front because my review isn't from someone the platform carried on the first try; it's from someone who plateaued hard and can tell you exactly which parts of e-GMAT broke the ceiling. Before committing, I tried and tested almost all other alternatives available and e-GMAT was the most holistic of them and by far the closest to the actual exam. That last part matters more than people realize. I took fifteen mocks on another platform the previous year and their verbal was nowhere near real-exam difficulty, which quietly miscalibrates you; e-GMAT's five mocks were the closest thing to the real test I found anywhere, and I treated them like dress rehearsals — same 10 AM slot, same café, same coffee, a full day of analysis after each one.
The structure is what turned "study harder" into "fix the specific leak." The path is simple: finish the module, then let the module questions expose your weaknesses, then diagnose rather than grind. The error log became the backbone of my prep — every mistake classified as execution error, process error, conceptual gap, or timing — and that classification is what told me, after my third 655, that my problem was no longer content at all. The quant question bank runs deliberately harder than the actual exam, and that's a feature: my weighted accuracy on it was around 85%, and on test day the real questions felt easy. That's how an engineer who "was always fine at math" converts fine into a Q90 — three times across my five attempts, in fact. Verbal was my weak section, and the platform's strategies (pre-thinking in CR before touching a single option) gave me a fixed, repeatable process for every question type; my weighted accuracy there reached 78%, and in DI around 73-74%. Walking into the exam, I wasn't hoping — I knew the data, and the data said I could perform.
The sectional mocks deserve their own mention. In the final stretch I stacked them back-to-back — verbal, then quant, then a 10-minute break, then DI — to simulate the real exam's flow and build stamina without burning a full mock. There are plenty of them on the platform, and they're the most underrated tool it has.
But the piece that actually got me from 655 to 675 was the Last Mile Push and my mentor, Shubham. After three identical scores there was nothing left to teach me about triangles or assumptions — what I needed was someone reading my data and my behavior from the outside. He's the one who told me to stop obsessing over verbal and let my strengths carry the score, to cut my study time from five-six sleep-deprived hours a day down to a hard maximum of three, and to stop putting my entire life's stakes on one sitting. Whenever I felt down in those last twenty days, I'd get on a call and come out lighter. That mentorship is the part of this course self-study can never replicate. If your concepts are done and the score still won't move — if you're the person rereading your third identical score report — this platform, and especially that program, is built for exactly you.