Start with the version you will sit

Before you plan your GMAT preparation, confirm the exam version you are taking. The GMAT is computer-based and adaptive. It assesses quantitative reasoning, verbal reasoning and data interpretation or analysis through multiple-choice sections; some versions also include an analytical writing task. The sections are time-limited. Do not assume that section names, order, timing or writing requirements are the same across versions. Check the official GMAT website for the current structure that applies to your booking.

The exam rewards reasoning from information in front of you, not simply recall. You need to interpret wording and constraints accurately, choose a workable method, and keep moving without letting a difficult item derail the rest of a section. Accuracy and pacing have to be practised together.

Begin by taking a diagnostic using materials appropriate to your version. Treat the result as a map, not a prediction. Record the question type, your choice, the time spent, and the reason for any error. Use these categories: knowledge gap, incorrect set-up, misread condition, weak elimination, calculation or reading slip, and time-pressure guess. A wrong answer is useful only if you can explain what caused it.

Build a map before building speed

Make a topic inventory from the official description and the practice materials for your version. Divide it into the actual question types you encounter, rather than broad labels such as “maths” or “verbal”. For each type, note the move that earns progress: translate a relationship into an equation, identify what a passage supports, compare answer choices against a definition, or read the relevant values from a table.

A MySummaries board on this topic ends up looking like this:

GMAT Reasoning, setup and pacingStudy
Quantitative · Verbal · Data · Writing where applicableReasoning, setup and pacing5 sections · 3 columns
Quantitative setup
  • Name the unknown before choosing an equation
  • Translate comparison language carefully: ‘x more than y’ means x = y + difference
  • Test units and permitted values before calculating
Elimination and checking
  • Reject an option for a specific contradiction, not because it sounds unfamiliar
  • Recheck signs, units and boundary conditions
  • If two options remain, compare each with the exact stem
Verbal evidence
  • Identify the question stem's task before judging options
  • For inference, require support from the passage; do not add outside assumptions
  • For argument analysis, separate evidence from the conclusion
Pacing review

Log time by question type in practice. Identify where accuracy falls when you speed up, and where careful work is taking longer than it needs to.

Data interpretation
  • Read labels, units, axes and footnotes before comparing values
  • Check whether the question asks for a value, relationship, or conclusion
  • Estimate first when it can rule out implausible options

Keep the board short enough to use during review. It is not a textbook. Attach your own worked examples, concise rules and error notes to the relevant sections. Mark a section as due when it has cards waiting, and update it when repeated practice reveals a better method.

Practise a repeatable reasoning routine

Use the same sequence across question types, adapting the middle steps to the task.

  1. Read the demand. Before solving, identify what the question is asking for. Is it a value, a comparison, a supported conclusion or an assessment of an argument?
  2. Set up. List the relevant facts, definitions, units and constraints. In a quantitative problem, define the unknown and translate relationships before calculating. In a verbal problem, locate the evidence and distinguish it from your own assumptions. In a data problem, check headings and units before interpreting a figure.
  3. Choose a method. Use the shortest valid route, not a trick that works only if your guess happens to be right. Estimation, substitution or elimination can be efficient when the options allow it; use exact calculation when the distinction requires it.
  4. Check the result. Ask whether the answer satisfies every condition, whether its size and units make sense, and whether the choice answers the stem rather than a nearby question.
  5. Move on deliberately. Set a personal pacing checkpoint during practice. If you cannot make progress, eliminate what you can and make a controlled decision rather than spending the time needed for several later questions.

For adaptive multiple-choice sections, consistency matters as well as individual accuracy. Avoid changing an answer without a reason you can name. Do not try to infer a hidden scoring rule from a short run of questions: practise the reasoning and pacing that the exam asks for, and check official guidance for score information.

A MySummaries Must not miss core for this board looks like this:

Must not miss coreReasoning, setup and pacing
State the question's demand and list the information that can answer it before working
Translate quantitative relationships with defined variables, units and constraints before calculating
For verbal questions, use the passage or argument evidence required by the stem; do not import unsupported assumptions
For data questions, inspect labels, units, axes and footnotes before drawing a conclusion
Estimate or sense-check when useful, then verify the selected option against the exact question

Train each section without guessing the format

Quantitative reasoning

Practise setting up problems before doing arithmetic. Write down what the variables represent, identify any restrictions, and translate the wording into a relationship. Then choose a calculation method. If the options are far apart, estimation may be enough; if they are close, calculate more precisely. Check the units and whether the answer is possible under the conditions given.

After each miss, solve the problem again without looking at your first working. Compare the two approaches. If the second attempt works, the issue may have been reading, set-up or execution rather than a missing concept. Keep a short note on the exact step that failed, such as reversing a comparison or overlooking a restriction.

Verbal reasoning

Read the stem to determine the task before evaluating choices. For a passage-based question, mark the claim or evidence relevant to that task. For an argument question, identify the conclusion and the support offered for it; then assess what an answer choice actually changes or establishes. Eliminate options by pointing to a contradiction, an unsupported leap or a mismatch with the question. “Sounds plausible” is not evidence.

Data interpretation and analysis

Start with the display, not the options. Read the title, labels, units, scale and any notes. Then identify the exact comparison or inference requested. Be alert to changes in denominators, time periods or categories: two values can look comparable but refer to different things. Use an estimate to check scale and direction before committing to a more detailed calculation.

Analytical writing, if included in your version

First verify on the official GMAT website whether your version includes a writing task and what it asks you to do. Practise organising a response around the task’s actual instructions. Make the reasoning explicit: identify the claim, assess the support provided, and explain what additional information would strengthen or weaken the conclusion where relevant. Reserve time to check that each paragraph supports your central analysis. Do not prepare by memorising generic claims that may not fit the prompt.

A MySummaries marked response can help you see how a written practice answer loses marks; the score and comments shown here are a platform assessment, not an official GMAT mark:

Paper — Reasoning, setup and pacing24:10
78%Reasoning, setup and pacing — marked7/9 marks · 24:10 taken

A retailer reports that stores displaying a new entrance sign had higher average sales than stores without it. The retailer concludes that installing the sign caused sales to rise. Analyse the reasoning and explain what information would help assess the conclusion.

7/9

The comparison does not establish that the sign caused the higher sales. Stores with the sign may differ in location, size, staffing or customer numbers. The retailer should compare similar stores over the same period and examine sales before and after installation. It would also help to know whether the stores were selected at random and whether other changes were made at the same time.

The response identifies the causal gap and proposes relevant comparisons. Its suggestions are clear, but it does not explain how customer numbers affect the interpretation of average sales or address the possibility that stronger stores were chosen first for the sign.

Missed

Explain whether the reported average is per store, per customer or total sales; these measure different outcomes

Connect non-random selection to the possibility that pre-existing store performance explains the difference

Model answerThe retailer has shown an association, not that the sign caused the increase. Stores chosen for the sign may already have had higher sales, or may differ in location, size, staffing or customer traffic. The comparison should define whether sales are totals or adjusted for customer numbers and compare similar stores over the same period. A randomised or otherwise well-matched comparison, together with sales before and after installation, would reduce the effect of pre-existing differences. The retailer should also report whether other changes occurred during the comparison period.

That is a MySummaries paper, filled with GMAT material. Yours is written from your own notes. Start free

Turn practice into a feedback loop

Use timed practice only after you understand the basic method for a question type. Start with a small set and review every answer, including correct guesses. For each question, write one sentence completing: “I chose this because…” If you cannot justify a correct choice, treat it as uncertain. If you can explain why a distractor is wrong, record the specific feature that rules it out.

Review errors in batches. Look for a pattern: perhaps you translate “at least” as “more than”, read the wrong row in a table, or accept a verbal option that is stronger than the evidence. Make a card for a single reusable fact or decision rule, not a whole question. Then revisit the problem later with the explanation hidden and solve it afresh.

Keep a pacing log for practice sets: question type, time taken, outcome and confidence. Compare accuracy at different speeds rather than simply trying to go faster. If accuracy drops sharply, the answer is not necessarily more speed work; it may be a more reliable set-up or an earlier decision to move on. If a type is accurate but slow, look for unnecessary calculations or repeated rereading.

A MySummaries lecture made from the board can turn one difficult process into a short listening review:

Lecture — Reasoning, setup and pacing8 min
Set up before you solveA spoken walkthrough of translating a quantitative relationship, checking its constraints and testing the result.
03:1608:02
Speed1×1.25×1.5×2×

Transcript · tap any word to jump there

Start with the question, not the arithmetic. Suppose a quantity is described as five greater than another quantity. Name the smaller quantity x; the larger one is x plus five. That is the relationship. Do not reverse it just because the sentence mentions the larger value first.

Now identify the constraint. If the problem says the quantities are positive integers, your choices must satisfy both the equation and that condition. A calculation that produces a decimal may be mathematically consistent with an equation but invalid for the problem.

Before selecting an option, substitute it into the original relationship. Check the sign, the units and the wording of the question. If the options are widely spaced, an estimate may help you reject several quickly; it does not excuse skipping the final check.

A workable weekly routine

Set the number of study sessions you can sustain, then give each session a defined job. For example, in a four-session week, use one session to learn or refresh a method, one for untimed questions and careful explanations, one for timed mixed practice, and one for error review and a short retest. If your version includes analytical writing, include a separate writing practice and feedback session. Adjust the balance to your diagnostic results rather than dividing time evenly among every topic.

At the start of each week, choose one priority from your error log. At the end, check whether the repeated error has decreased in fresh practice. If not, break it down further: was the rule unclear, the stem misread, or the decision too slow? Increase the difficulty of timed work only when you can explain the method under untimed conditions. Leave enough time near your exam to practise complete sections under the applicable version’s conditions, without inventing a schedule from another version.

How MySummaries helps

Build a revision board from your own GMAT notes and practice explanations, then use it to generate question-specific flashcards, written practice where relevant, and examiner-voice audio lectures. You can review recurring errors and focus the next session on the reasoning steps that need work. Start at portal.mysummaries.app.