The campus aptitude round, and how to actually prepare for it

On 6 September 2026 I opened faceprep’s pattern pages for the Deloitte and CoCubes campus aptitude tests as part of aptitude test preparation research, to check whether the section timing is as tight as candidates describe. Deloitte’s version runs 75 questions across 95 minutes: 25 quant in 35 minutes, 25 logical reasoning in 35 minutes, 25 verbal in 25 minutes. Do the arithmetic and a quant question gets 84 seconds. Verbal gets 60. Neither section has negative marking, according to the same source, and neither does CoCubes or TCS NQT’s 2026 pattern, which I checked separately. That absence of a penalty for wrong answers changes the whole strategy, and most study advice online is still written for a penalty that most current tests don’t actually apply.

This round exists to filter, not to evaluate you fairly. A company hiring 4,000 graduates off campus drives cannot read 4,000 resumes closely. The aptitude test is the cheap first cut before anyone looks at your code, your project, or your GPA. Most candidates lose here, not in the interview.

What aptitude test preparation is actually preparing you for

Four sections show up across nearly every company pattern, in different proportions. Quantitative aptitude covers percentages, ratios, profit and loss, time and work, time-speed-distance, and number systems, the arithmetic-heavy topics from school math rather than anything requiring calculus. Logical reasoning covers coding-decoding, series completion, blood relations, syllogisms, seating arrangements, and puzzles that reward pattern recognition over calculation speed. Verbal ability tests reading comprehension, sentence correction, para-jumbles, and synonym-antonym recall, closer to a school English exam than a literature course. Data interpretation, sometimes folded into quant and sometimes its own section, hands you a table or a chart and asks several questions off the same dataset, so getting the first read of the chart wrong compounds across every question tied to it.

Weighting differs by company. Deloitte splits its three sections close to evenly by question count but not by time, since verbal gets 25 fewer minutes for the same 25 questions. TCS NQT runs a longer, two-part structure: a 75-minute Foundation part with 65 questions across numerical, verbal and reasoning, followed by a 115-minute Advanced part that layers in coding. Knowing your target company’s specific split matters more than knowing the generic aptitude syllabus, because a candidate who over-prepares verbal for a quant-heavy pattern has wasted real hours.

The time budget is the actual difficulty

None of these questions are individually hard. A profit-and-loss problem that would take four unhurried minutes to work through carefully has to be solved, checked, and moved past in under 90 seconds on test day. That compression is where most candidates fail, not on the math itself.

Two skills close that gap, and neither is “study more formulas.” The first is recognizing question types fast enough to pick a solving method before you’ve fully read the question twice. The second is knowing when to abandon a question. A 90-second cap means any question you don’t recognize within 20 seconds should get a guess and a skip, not a fight. Candidates who insist on solving every question in order, in full, tend to run out of time on the last eight questions of a section. Several of those late questions are usually easier than the ones they got stuck on earlier.

Accuracy versus attempts, without a penalty

Here’s the part that trips people up specifically because most current tests, Deloitte’s and TCS NQT’s included based on what I checked, carry no negative marking. Under that rule, a blank answer and a wrong answer score identically: zero. Leaving a question blank has no advantage over guessing, so guess on everything you don’t finish, always, with no exception. That’s a mechanical rule, not a judgment call.

Where accuracy still matters, even without a penalty, is the adaptive or sectional-cutoff design some platforms use. AMCAT, for instance, adjusts question difficulty based on your running accuracy and scores you partly on the difficulty band you sustained, not just raw correct count. A candidate answering 12 of 16 questions correctly at a harder difficulty tier can outscore one answering 14 of 16 at an easier tier. That’s a different game from a flat-scoring test, and it means slowing down slightly to stay accurate on an adaptive platform can beat racing through easy questions. Check which model your target company uses before deciding your pace.

Where negative marking does still appear, mainly in aptitude-heavy exams outside pure campus drives such as CAT and some public-sector recruitment tests, the calculation flips. A wrong answer there costs a fraction of a right answer’s value, commonly a third, so guessing blind on a question you have zero insight into is a losing bet on expected value, while guessing after eliminating two of four options usually isn’t. Know which rule governs your specific test before you walk in. Guessing strategy under a penalty and guessing strategy under no penalty are opposite advice, and applying the wrong one costs real marks.

A 3-week aptitude test preparation plan that fits around classes

Three weeks of aptitude test preparation is enough if you’re consistent, and useless if you cram it into the last four days, which is what most people actually do. Here’s a version built around roughly an hour a day, more on weekends.

Week one, diagnose and drill fundamentals. Take one full timed mock on day one before studying anything, purely to see your real starting point and your current per-section timing. Spend the rest of the week on quant fundamentals: percentages, ratios, averages, profit and loss, time and work. These five topics alone cover a disproportionate share of quant questions across most company patterns, so mastery here pays back faster than spreading thin across the full syllabus.

Week two, reasoning and verbal, plus a second mock. Reasoning topics reward pattern drilling: do 20 to 30 questions a day across coding-decoding, series, syllogisms, and seating arrangements until the pattern recognition becomes close to automatic. Verbal needs less raw practice and more reading habit: 15 minutes of dense, unfamiliar text daily (editorial pages work well) does more for reading comprehension speed than grammar drills alone. Take mock two at the end of the week and compare your section timing against mock one specifically, not just your total score.

Week three, full mocks and weak-topic triage. Two to three full timed mocks this week, reviewing every wrong answer for the reason it went wrong: a concept gap, a careless error, or a time-pressure guess. Concept gaps get one more focused study session. Careless errors get a checklist you actually read before submitting each section. Time-pressure guesses tell you where to adjust your skip threshold. By the final mock, your accuracy should matter less to you than whether your pacing holds up across all three sections without a late collapse.

Where the return on an hour of study is highest

Not every topic deserves equal time. Percentages, ratios and time-and-work show up across almost every company’s quant section and have a small, learnable set of question patterns, so they return the most score per hour studied. Number series and coding-decoding in reasoning are similarly high-return: a handful of pattern types cover most of what appears. Reading comprehension is the opposite. It’s high-volume in most patterns but slow to improve through short-term drilling, since it depends on baseline reading speed built over years, not weeks. If your three weeks are limited, spend them on quant fundamentals and reasoning patterns first, and treat verbal as maintenance rather than a growth area unless you’re starting from a genuinely weak base.

And if you’re already clearing 80 percent on quant mocks by week two, stop drilling quant. Put that extra hour into a fifth full mock instead. Pacing across all three sections under real time pressure is the harder skill by then, not another percentages worksheet.

Data interpretation rewards a specific habit more than raw practice volume: read the chart’s title, units and any footnote before touching the first question. A wrong assumption about whether a table shows thousands or lakhs invalidates every question built on it, and that single habit prevents more lost marks than an extra week of DI drilling would.

What this looks like once you clear it

Clearing the aptitude round gets you to the round that actually resembles the job: a coding test, a group discussion, or a technical interview, depending on the company. If your target is a company running a structured online assessment beyond just aptitude, our breakdown of the Amazon online assessment covers what changes once coding and work-style sections get added to the mix. For the fresher-specific version of the salary conversation that follows an offer, our salary expectations guide covers what to say when a recruiter asks first.

If you’re running mock interviews to prepare for what comes after the aptitude round clears, our interview copilot gives live, discreet guidance during practice and real sessions alike, and the free tier covers 15 credits a month if you want to try it before a placement season gets busy.

One thing I’m not certain about

I don’t have solid data on how much aptitude-round performance actually predicts on-the-job success, and I’d guess the correlation is weaker than the amount of hiring weight placed on it. Companies keep using these tests anyway because they scale to thousands of candidates in a way that interviews don’t, not necessarily because they’re the best predictor available. That’s a structural reason to take the round seriously regardless of whether you think it’s fair. It decides whether anyone reads the rest of your application.