First 30 days of GMAT prep

The first month of GMAT preparation decides more about a candidate's eventual scaled score than almost any other stretch of the calendar. Most candidates who arrive in week five having "studied for 20 hours" still cannot tell a Data Sufficiency stem from a Two-Part Analysis prompt, and the work they actually need to do is rebuild habits, not pile on hours. This article is a tutor's view of how to spend the opening thirty days: which diagnostic to run on day one, which question types to touch in week one, when to introduce timed drills, and what to leave untouched until the second month. The advice below is written for working candidates, but the structure works equally well for full-time students who want to avoid the classic mistake of reading the entire Official Guide before answering a single question.

Day 1 to day 3: establish a clean baseline before opening a study book

Every serious GMAT (formerly the GMAT Focus Edition) plan begins with a baseline, and the baseline must be untimed and unbroken. Sit one full Quant section and one full Verbal section under relaxed conditions, in a quiet room, with a notebook open beside you. The point is not to score well. The point is to see the GMAT format with fresh eyes and to record the gap between what the question was asking and what you actually answered. A candidate who has never seen Critical Reasoning will guess; a candidate who has never seen Two-Part Analysis will misread the response format entirely. Both behaviours are correct on day one and intolerable by day thirty.

After the baseline, write down three numbers. First, the time you spent on the section. Second, the number of items you would mark as "I had a real method, I just made an arithmetic slip" versus "I had no method at all." Third, the items where you talked yourself out of the right answer. These three counts form the spine of your first month. In my experience, the third number is the largest and the most fixable. Candidates routinely talk themselves out of correct answers because the GMAT phrased a condition in the negative or used a word like least in a way that conflicted with their first read. Writing the counts down turns vague anxiety into something you can schedule against.

Resist the temptation in these first three days to binge content. A 30-minute review of the official question-type taxonomy is plenty. The GMAT has a small, finite menu: in Quant, Problem Solving and Data Sufficiency; in Verbal, Reading Comprehension, Critical Reasoning, and now a leaner Two-Part Analysis replacing the old Sentence Correction family; in Data Insights, the multi-source set, Graphics Interpretation, Two-Part Analysis, Table Analysis, Data Sufficiency, and Identifying Relevant Information. Knowing the names is the entire goal of day one to day three. Deep familiarity with each item family comes later.

• Day 1: one untimed Quant section, log the three counts above.
• Day 2: one untimed Verbal section, same log, then a 20-minute read of the official question-type list.
• Day 3: light review of the logs, no new questions, decide on a fixed daily study window.

End the third day with a written commitment to a study window. For most working candidates this is 60 to 90 minutes on weekdays and a longer block on one weekend day. The window must be small enough to survive a bad week at the office, or the plan collapses by day twelve.

Week 1: meet each question family once, untimed, on paper

The job of week one is recognition, not performance. For each of the roughly seven item families in the GMAT, the candidate should solve three to five questions, read the official explanation, and write one sentence about the underlying skill. Three sentences for Data Sufficiency, since that item family carries the heaviest weighting on Quant. A schedule that works for a working candidate is one item family per weekday evening, 45 minutes, with the longer Saturday block reserved for Reading Comprehension passages, which cannot be triaged in 45 minutes.

For Quant, this means seeing Problem Solving in its algebra, arithmetic, word-problem, and geometry variants. Most candidates discover on day four that they are fast on algebra and slow on rate problems, or confident on integer properties and shaky on probability. Neither of those discoveries is news; the news is that the discovery happens on day four, not on day twenty-five under timed pressure. For Data Sufficiency, the recognition task is to internalise the five decision rules: the statement must be sufficient on its own, the two statements together, both must be necessary, no extraneous information counts, and an answer is unique. A candidate who can recite those five rules by day seven has bought themselves ten points of slack later in the plan.

For Verbal, week one is the time to read one Critical Reasoning argument skeleton per day. The skeleton is: conclusion, premise, assumption, strengthen or weaken target. Most candidates skip the skeleton step and dive into answer choices, which is why they pick the choice that sounds smart rather than the choice that addresses the gap. Reading Comprehension in week one is about passage mapping: read a 250-word passage once, write a one-line topic, write a one-line scope, write a one-line author stance. If a candidate cannot produce all three in under three minutes, the question will beat them in two.

Data Insights is the section most candidates underestimate. The seven item families do not all reward the same reading habit. Graphics Interpretation punishes axis-skimmers; Table Analysis punishes column-skimmers; Multi-Source Reasoning punishes tab-skimmers. Spend week one deliberately slow on each family, and note the unique reading habit each one demands.

Common pitfalls and how to avoid them in week 1

The most common week-one mistake is buying six prep books. Pick one source for question banks, one for explanations, and stop. The second most common mistake is timing the first three questions. Untimed practice in week one teaches the right things; timed practice in week one teaches panic. The third is treating Data Sufficiency as arithmetic. Data Sufficiency is a categorisation problem with five answer choices; if a candidate is reaching for a calculator, the categorisation step has been skipped.

Week 2: build a per-item-family error log and an untimed accuracy target

By week two, the candidate has seen every item family at least once. The job of week two is to convert recognition into a method, and to convert a method into a measurable accuracy. The mechanism is an error log. After every practice set, the candidate writes the item family, the question topic (rates, integers, strengthen-the-argument, axis label, third-tab inference), and a one-line reason the wrong answer was wrong. A log that fits on one side of A4 per day is the right size. A 30-page spreadsheet will be abandoned by day twenty.

Set a target accuracy for each family by the end of week two. The targets are deliberately imperfect: 60% on Data Sufficiency, 65% on Problem Solving, 70% on Critical Reasoning, 60% on the slower Data Insights families. The numbers are not a measure of talent; they are a forcing function. If a candidate's actual accuracy on Critical Reasoning is 45% at the end of week two, the plan for week three is obvious, and the candidate will not waste time drilling the families they are already good at.

Untimed accuracy is the right metric for week two. Time enters the plan in week three. A candidate who tries to time themselves in week two will trade accuracy for speed, which is the wrong trade at this stage. The GMAT is, in scoring terms, a test of accuracy under time pressure, and the only way to train that is to master accuracy first and then layer time on top.

Reading Comprehension deserves its own paragraph. In week two, the candidate should read one full passage per day and answer all associated questions, then mark only the questions where the wrong answer choice was attractive. Most candidates discover that their RC errors are not comprehension errors at all; they are misread errors, where they substituted a word in the question stem (for example, most strongly supports read as most directly challenges) and chose an answer that was actually defensible against the misread. The fix is to underline the operative verb in the question stem before reading the choices. This is a 10-second habit that pays for itself many times over the rest of the plan.

Week 3: introduce time pressure on the families you have already mastered

Week three is the first time the candidate should reach for a timer. The principle is selective: time only the families where the candidate hit their week-two accuracy target, and leave the weaker families on untimed practice. The reason is tactical. Time pressure is a magnifier. If a candidate is at 70% accuracy on Critical Reasoning untimed, layering 90 seconds per question will drop them to perhaps 60%, which is still a strong base. If a candidate is at 40% accuracy on Data Sufficiency untimed, layering time will drop them to 25%, which produces nothing useful and burns confidence.

A clean week-three cadence is four timed sets of 10 questions on the strong families, each set followed by 20 minutes of error-log review. That is roughly 90 minutes on a weekday, plus the longer Saturday block. The Saturday block in week three should be a single full Verbal section under timed conditions, which is the first time the candidate sees a 45-minute window of mixed question types. Most candidates feel rushed on the first timed Verbal section; the feeling is informational, not a verdict. The data point that matters is which question they rushed, and why.

For Data Insights, week three is also when the candidate should practise the mechanics of the on-screen calculator and the multi-tab interface. These are software skills, not reasoning skills, and most candidates under-train them. Twenty minutes of clicking between tabs, dragging rows, and entering expressions is enough. The reason this matters is that a candidate who fumbles the interface in week three loses perhaps 30 seconds per multi-source set. Across five multi-source sets in a real test, that is two and a half minutes of recovered thinking time, which is the difference between attempting and skipping the final Two-Part Analysis.

Common pitfalls and how to avoid them in week 3

Do not time every question family at once. Pick two strong families, time those, leave the rest on untimed practice. Do not interrupt the error log to start a new question set. The error log is the single most leveraged habit in the entire 30 days. Do not begin week three without a quiet 30-minute review of the week-two log; the patterns in that log dictate the schedule for week three.

Week 4: take the first full-length baseline under real conditions

Week four is the integration phase. The candidate has one item family per weekday where they are now reliably accurate, an error log with three weeks of patterns, and software skills for the Data Insights interface. The job of week four is to take a full GMAT practice test, end to end, under realistic conditions. This is not the real test. It is a diagnostic for the second month.

Realistic conditions mean: same start time as the planned real test, a quiet room, a single break policy, no phone, no notes, no error log during the test. The candidate records the section-level scaled scores, the time spent on each section, and a single overall feeling ("ran out of time on Quant, finished Verbal with minutes to spare"). The numeric output is the second month's input.

After the baseline, the candidate writes a one-page plan for the next thirty days. The plan should answer four questions. Which two Quant families are weakest, and what specific method needs to be installed? Which two Verbal families are weakest, and what reading habit needs to change? Which Data Insights family consistently eats the most time, and what is the triage rule? What is the next baseline date, and what scaled-score improvement would make the plan worth continuing?

The first full-length baseline will almost always feel worse than the candidate expects. That is normal. A candidate who has only ever practised untimed question families will, on the first timed integrated test, drop several points below their per-family accuracy. The drop is the cost of integration, not a verdict on the plan. The second full-length, taken in week six or seven, is the first one that is genuinely diagnostic of the candidate's score ceiling.

Question-type priority order for the first 30 days

Not every question type carries the same weight in the first 30 days, and a working candidate cannot drill everything at once. The priority order below is what I would hand a candidate walking in cold on day one, and it deliberately leaves the lower-priority families to month two.

The table is a sequencing tool, not a ranking of importance. Two-Part Analysis is more important than Reading Comprehension in terms of test design; it sits at priority nine here only because a working candidate cannot drill nine families in 30 days, and Two-Part Analysis rewards a transfer of habit from the other Data Insights families rather than a fresh method.

Daily and weekly cadence for a working candidate

The plan above only works if it survives a bad week at the office. The cadence below is what I would hand to a candidate in a 50-hour-a-week job, and it deliberately leaves the weekend for the heavier work.

• Monday: 60-minute untimed set on the week's weakest Quant family, error log, no timer.
• Tuesday: 60-minute untimed set on the week's weakest Verbal family, error log, no timer.
• Wednesday: 60-minute timed set on a strong family from week two, error log, light review.
• Thursday: 60-minute Data Insights practice focused on the interface (tabs, drag, sort) plus one item family.
• Friday: 45-minute review of the week's error log, no new questions, write the one-sentence plan for the weekend.
• Saturday: 2.5-hour block: one full Verbal section timed, plus a deep review of the worst five errors from the week.
• Sunday: Rest, or a 60-minute light read of one Reading Comprehension passage and a long walk.

That is roughly 8 to 9 hours per week, which is sustainable for most working candidates and dense enough to produce a measurable change by the end of month one. The Sunday rest is non-negotiable. Burnout by day twenty-five is the most common cause of plan abandonment, and a single rest day is the cheapest insurance against it.

What to leave out of the first 30 days

A working candidate in month one has roughly 35 hours of total study time, and the temptation to spend it on advanced topics is strong. The list below is what to deliberately defer.

First, advanced quant content. Permutation gymnastics, probability trees with conditional setups, and geometry proofs with multiple inscribed shapes are all real GMAT content, but they appear as one or two questions per section. A candidate who spends 10 hours on advanced combinatorics in week three has spent 10 hours that should have gone into Data Sufficiency decision rules. The advanced content belongs in month two, when the candidate has the accuracy baseline to know whether they need it.

Second, full-length retakes. The first full-length is the diagnostic; the second should be at least 14 days later, in week six or seven, when there is enough new data to act on. Retaking the full-length in week four is a confidence trap, not a study method.

Third, the official practice question banks. These are best used in month two and month three, after the candidate has a per-family method and an error log. Wasting the official bank in week one is a real cost; the questions are finite and the patterns are subtle.

Fourth, admissions-side planning. Application deadlines, school research, and recommendation letters all matter, and none of them belong in the first 30 days. A candidate who lets the application calendar leak into the study window is the same candidate who arrives in month two with no error log and no accuracy data. The two plans should run in parallel and never overlap.

Conclusion and next steps

The first 30 days of GMAT preparation are about building the substrate on which month two and month three will run. A clean baseline on day one, one item family per weekday in week one, an error log and accuracy targets in week two, selective timed practice in week three, and a single integrated full-length in week four. A working candidate who executes that cadence with 8 to 9 hours per week arrives in month two with a written plan, a real method on the strongest families, and a measurable accuracy floor. That is the only honest starting point for the score climb that follows.

For candidates building that opening cadence, a focused diagnostic on the first-week item-family recognition is the natural starting point, since the rest of the 30 days falls out of the patterns the diagnostic surfaces.

Frequently asked questions