Start by attempting one previous-year CAT paper without a timer to see where you actually stand — most beginners are surprised by how different CAT is from college-level maths and English. The core logic of how to prepare for the CAT exam is the same for everyone: build concepts section by section (QA, VARC, DILR), move to sectional tests, and then shift most of your time to full-length mocks with honest analysis. Keep resources minimal — one book or course per section — and fix a non-negotiable daily slot of two to three hours. If you are unsure how to start CAT preparation at all, treat your first 30 days as concept-building plus one diagnostic mock, not advanced problem sets.

RC in CAT preparation refers to Reading Comprehension, the passage-based part of the VARC section. In recent CAT papers, roughly four passages with four questions each — around two-thirds of the VARC section — come from RC, making it the single highest-weight area. The passages are usually dense and opinion-based (philosophy, economics, science, culture), and you are tested on the author's main idea, tone, and inferences rather than vocabulary. Daily reading of editorials or long-form essays, combined with regular timed passage practice, is the most reliable way to score well here.

You only need one trusted book per section — a standard Quantitative Ability book like Arun Sharma, a DILR practice book, and previous-year CAT papers for VARC. Most aspirants waste weeks collecting CAT preparation books instead of revising and solving the two or three they already own. From the third month onwards, mock tests and past papers should drive your preparation, with books used only for concept refreshers and drills.

For a disciplined aspirant, CAT preparation online is easily as effective as classroom coaching — online test series, video solutions, and doubt-solving communities now cover everything a physical batch offers, usually at a lower cost and with zero commute. Classroom coaching mainly helps if you need fixed timings, external structure, and peer pressure to stay consistent. Choose online if you are self-motivated or a working professional; consider a classroom only if you struggle to follow a routine on your own.

For most aspirants, a CAT preparation duration of six to nine months works well — long enough to cover all three sections and write 25–30 mocks, short enough to avoid burnout. Working professionals usually need the higher end of that range, planning around two focused hours on weekdays and longer sectionals on weekends. Starting between January and March for the November exam gives comfortable runway; if you start after June, you will need to trim depth and rely more heavily on mock analysis. Consistency across months matters far more than the total number of study hours.

GDPI stands for Group Discussion and Personal Interview — the second stage of MBA admissions in India that follows your written-exam shortlist. Some institutes run a group discussion or case discussion, while others use a Written Ability Test (WAT) in its place, but every version ends with a personal interview covering your academics, work experience, hobbies, and reasons for pursuing an MBA. A strong CAT or XAT score only gets you to this stage; the final admission offer depends heavily on how you perform in GDPI.

CAT preparation is skill-based — you grind quant concepts, reading speed, and puzzle accuracy — while GDPI preparation for MBA is profile-based, testing your self-awareness, communication, general awareness, and clarity of goals. You cannot clear GDPI by solving question banks; instead, you prepare by reading the news daily, writing and refining answers to questions like "tell me about yourself" and "why MBA," and rehearsing in mock group discussions and interviews. The timelines also differ: CAT prep runs for months before the exam, but serious GDPI prep should begin as soon as you expect a shortlist, since interview calls often come within weeks of results.

A GDPI preparation course is helpful but not compulsory — plenty of candidates convert IIM and other top B-school calls using peer mock interviews, daily news reading, and honest self-review. Structured help pays off mainly in feedback: catching vague answers, filler speech, and gaps in your story that you cannot spot yourself. If your communication is already strong and you have friends to mock with, self-preparation works; if you ramble under pressure or your profile has tricky areas like low academics or a career switch, a few guided mocks are worth the money.

IIM interviews are tough mainly because they are unpredictable, not because panels are trying to trap you. Most questions come from your own form — graduation subjects, work experience, projects, hobbies, hometown, and current affairs — so a candidate who knows their own story in depth usually finds the interview conversational. Stress-style panels do exist at some institutes, but across IIMs the interview is generally a 15–25 minute test of clarity, honesty, and how gracefully you handle "I don't know." Four or five mock interviews before your first real call makes a significant difference.

Real IIM mock interview questions come from three places: past interview experiences shared by candidates on forums and YouTube, question banks compiled by coaching institutes, and — most importantly — a mock panel that builds questions from your own application form. A large share of an actual IIM interview is generated from what you write on your form, so being grilled on your specific academics, work experience, and hobbies is more realistic than practising from a generic list. Combine a broad set of IIM mock interview questions with at least one personalised mock where the interviewer has read your full profile.

An IIM mock interview course is worth it if you have no access to seniors or alumni who can take your mocks — it gives you panel-style pressure, profile-specific feedback, and enough repetitions (ideally four to six mocks) to normalise the interview environment. If you already have IIM alumni in your network or a peer group with calls in hand, free mock exchanges can work nearly as well. The real value lies in personalised correction, so judge any course by who conducts your mock and how detailed the feedback is, not by the number of recorded lecture hours.

Start with a real problem, not the technology: pick one narrow use case, talk to potential users, and only then choose your stack. You can go surprisingly far today with LLM APIs and no-code or low-code tools before writing serious code, so the fastest path is to prototype a small MVP, put it in front of real users, and iterate on actual feedback. Focus on learning prompting, basic API integration, and how to evaluate model outputs — that combination covers most early product decisions. Ship small and early, because most AI products fail due to a poorly chosen problem, not a weak model.