GMAT Focus exam week is the narrow window between the end of structured study and the moment you sit down at the testing centre or open the online proctor window. Everything you do in those seven days shapes the score you walk away with, because the content review is essentially finished and the only variables still in play are fatigue, pacing rhythm, anxiety, and the small logistical mistakes that can derail an otherwise well-prepared candidate. This article lays out a tutor-grade plan for the 168 hours before a GMAT Focus sitting, with a daily block structure, a mock-test placement decision, a test-morning protocol, and a list of common pitfalls that quietly cap scores in the high 70s.
Two features of the GMAT Focus format make exam-week planning different from older fixed-order sittings. First, the section timing is tighter: each of Quant, Verbal, and Data Insights is a 45-minute block with no unscored experimental items, so every minute of the sitting counts toward the scaled score. Second, the adaptive engine locks the section order — Quant, Verbal, Data Insights — and there is no break between Verbal and Data Insights, only an optional 10-minute break after the first section you choose to take it after. Both features change how the final week should be sequenced, and the plan below is built around them rather than around generic pre-test advice.
The shape of the seven days: a high-level skeleton
Think of the final 168 hours as four phases, each with a single job. Phase one, days seven through five, is taper — you stop adding new content and let the patterns you already own consolidate. Phase two, days four and three, is mock-test and review. Phase three, days two and one, is recovery and logistics. Phase four, test morning, is execution. The mistake I see most often is candidates treating the whole week as a single uniform block of grinding, which is exactly the wrong shape. Your brain does not absorb anything new in the 48 hours before a sitting; it consolidates what it has already absorbed, and it does that best when sleep, food, and movement are stable.
A useful mental model is the athletic taper. A runner does not run a hard session the day before a race. A pianist does not learn a new piece the night before a recital. The GMAT Focus is a performance under time pressure, and the cognitive load of the sitting is real, so the final week is structured so that you arrive at the testing chair rested, fluent in the question families, and clear on your pacing budget. The block structure below assumes a Sunday test day; if your sitting falls on a different day, shift the days forward or backward by the same offsets rather than reorganising the logic.
Daily time budget across the week
For most candidates, the right total study time across the seven days is between 8 and 12 hours, not 25. A typical distribution looks like this: day seven, 90 minutes of light mixed practice; day six, 60 minutes of weak-area drill only; day five, off or 30 minutes of untimed warm-up; day four, one full mock test (about 2 hours 15 minutes including the optional break) plus 60 minutes of review; day three, 90 minutes of error-log review and pacing drills; day two, off or a 30-minute light walk-through of the official starter questions; day one, off except for logistics; test morning, the warm-up protocol described later in this article. The total sits between 8 and 10 hours for most readers, which is plenty because the marginal hour at this stage is almost always more useful as sleep than as practice.
Day-by-day block plan for the final week
Day seven: light mixed practice, no new content
Spend 90 minutes on a mixed set pulled from the question families you have already trained. For Quant, rotate 8 to 10 questions across Problem Solving and Data Sufficiency, with a hard cap of 2 minutes 15 seconds per question including the time to read the stem carefully. For Verbal, rotate 6 to 8 questions across Reading Comprehension, Critical Reasoning, and the one-question-per-pass items, timed at 1 minute 45 seconds. For Data Insights, do one each of Multi-Source Reasoning, Table Analysis, Graphics Interpretation, Two-Part Analysis, and Data Sufficiency, timed at the per-question budgets you trained with earlier in the cycle. The point of this session is not to learn anything new; it is to keep your pacing hand warm and to confirm that the question families still feel familiar after a day or two away from the screen.
End the session with a 10-minute error walk-through. You are not re-solving errors from scratch; you are scanning your error log for the two or three patterns that have shown up most often, and you are writing one sentence each on what the trap looked like and what the correct response was. The sentence itself is the consolidation tool — verbalising the pattern, even in a muttered aside to yourself, is what locks it into retrieval-ready form. Do not start a new topic. Do not read a new chapter of a prep book. The day before the day before a tapered review is not a place for new material.
Day six: weak-area drill, no full sets
Drill only the question family that has been your lowest-yielding pattern in the last two weeks of practice. If that is Two-Part Analysis in Data Insights, do 6 to 8 Two-Part stems back to back, with explicit focus on the column-entry mechanic and the shared-prompt logic. If it is Data Sufficiency, do 6 to 8 stems with a hard rule: you will not solve for the answer, you will only test the two statements against the question stem and choose an option. If it is Critical Reasoning inference questions, do 6 to 8 stems with the same rule. The point of the day is single-pattern sharpening, not breadth. The brain consolidates one pattern much more efficiently than three, and you have a hard ceiling of about 60 minutes for this block.
Stop at 60 minutes no matter how the drill is going. If the pattern clicks, stop. If the pattern is still fuzzy, stop anyway and sleep on it — sleep is the consolidation engine, and extra drilling at this stage tends to introduce a new error mode (over-reading, second-guessing) rather than reinforcing the trained response. The day-six block also doubles as your last opportunity to confirm that any accommodations, whiteboard supplies, or allowed water bottle are still in working order, but keep that logistics check under five minutes so it does not bleed into anxiety territory.
Day five: full rest or a 30-minute untimed walk-through
Most candidates are better off taking the day fully off. If you are the type who gets restless without a small amount of practice, do no more than 30 minutes of untimed mixed items from the official starter set, with the timer visible but ignored. The purpose of the day is to let the previous two sessions of mixed practice and weak-area drill settle. The cognitive science is straightforward: the hippocampus replays recently learned material during slow-wave sleep, and that replay is what moves patterns from short-term to long-term memory. You cannot force the replay by drilling harder; you let it happen by resting.
Use the day to do the things that are not test prep but support test performance: movement, hydration, social contact, daylight exposure, and a normal sleep schedule. Candidates who sit the GMAT Focus after a week of all-nighters and skipped meals are not faster or sharper; they are simply running a sleep and glucose deficit that the adaptive engine will punish. The 30-minute untimed walk-through exists only for the small number of candidates who find total rest anxiety-provoking — and even then, it is capped at 30 minutes and done with the explicit instruction to yourself that the session is not scored and not diagnostic.
Day four: the mock-test decision
Day four is the most contested day of the week, and the answer depends on where you are in your overall prep arc. The rule of thumb I use with my own candidates is this: if you have taken at least three full-length adaptive mocks earlier in the cycle and your last two mock scores are within 10 points of each other, take one more mock on day four. If you have taken fewer than three mocks, or if your last two mocks were more than 20 points apart, skip the day-four mock and use the time for review and pacing drills instead. A mock that produces a new low score three days before the real test is worse than no mock at all, because it injects anxiety at the worst possible moment.
If you do sit a day-four mock, the review block is at least as long as the mock itself. Spend 60 to 90 minutes walking through every missed question, every question you got right but flagged, and at least three questions you got right and did not flag, to confirm the right answer was not a lucky guess. The review is where the score actually moves. End the day by 6 pm if possible; the mock is cognitively expensive and you want evening free for recovery. Do not analyse a mock at 11 pm the night before review day — your sleep is worth more than one more re-read of a question stem.
Day three: error log and pacing drills
The day-three block is the consolidation session. Open your error log, sort the entries by question family, and pick the three families that have the most entries. For each family, do 4 to 6 questions of that type with a strict pacing budget — 2 minutes 15 seconds for Quant, 1 minute 45 seconds for Verbal, and the family-specific budget for Data Insights. The aim is fluency under the clock, not a new personal best. If a question is going long, mark it, move on, and return only if there are 30 seconds left in the budget.
End the block with a 10-minute scan of your pacing chart. If you are timing each section, you should be able to draw, from memory, the question number at which you banked a 30-second buffer on the last three practice sections. That number is your pacing anchor. On test day, when the timing gets tight around question 8 or 9, you will know whether you are still inside the buffer or already outside it, and that single piece of information is what lets you decide between solving and triage. The day-three block is also a good moment to re-read the official test-day rules once, slowly, so the procedural details are not new on test morning.
Day two: full rest, logistics only
Day two is for the logistics checklist and nothing else. Confirm the test centre address or the proctoring app installation. Confirm the ID meets the photo and name-match requirements. Confirm the time zone of the appointment — candidates sitting in a different time zone from where they trained frequently misread the start time and arrive late or too early. Lay out the clothes, the snack, the water, and any permitted items the night before, not the morning of, because decision fatigue on test morning is real and you want every saved minute to go toward warm-up, not toward choosing a snack.
Eat a normal dinner. Avoid alcohol. Avoid a heavy late meal. Aim for a bedtime that gives you at least 8 hours before the alarm, even if that means an earlier-than-usual night; the GMAT Focus is a sleep-sensitive task, and the difference between 6 and 8 hours the night before is visible in pacing on the harder Quant stems. If you find yourself unable to sleep, do not lie in bed with the lights off for two hours — get up, do 20 minutes of a boring non-screen activity, and return to bed. Sleep quality matters more than sleep quantity on a single night, but both are easier to protect on day two than to recover on day one.
Day one: the lightest possible touch
Day one is the day before the test. The only GMAT-related activity permitted is a 15 to 20 minute warm-up in the late afternoon or early evening: 5 Quant items, 5 Verbal items, 4 Data Insights items, all easy or medium, all timed, all for the purpose of getting your hand back on the keyboard and your eye back on the stem format. Do not do hard items. Do not start a new pattern. Do not solve a question you have not seen before. The aim is motor warm-up, not learning.
Spend the rest of the day on the things that are not test prep. Walk. Eat at your normal times. Drink water. Talk to a human. Watch a film. Avoid the temptation to read Reddit threads about the GMAT Focus at 11 pm — the marginal information you could pick up is not worth the anxiety it can introduce. Set two alarms, the second one 10 minutes after the first, and put your phone across the room so you have to stand up to turn it off. The morning protocol begins the moment the second alarm goes off.
Test-morning protocol: the 90 minutes before the timer starts
Test morning is the part of the week most candidates under-plan, and it is the part where the small mistakes compound. The first 90 minutes look like this. Minute 0 to 10: out of bed, water, light stretch, no screens. Minute 10 to 25: a breakfast with protein, slow-release carbohydrate, and minimal sugar — eggs and toast, oatmeal and fruit, yoghurt and granola. Avoid a large coffee on an empty stomach; the caffeine curve is sharper than you want it to be in the second 45-minute block. Minute 25 to 50: a short walk or light movement, ideally outside, to bring the cortisol rhythm up gently. Minute 50 to 75: travel to the centre or sign in to the proctoring app, with the ID and confirmation number in hand.
Minute 75 to 90 is the only test-prep activity of the morning. Do 8 to 10 easy mixed questions, untimed, on the official starter set or a paper print-out of familiar stems. The point is to wake up the pattern-recognition system, not to learn anything. If a question feels slow, skip it. If a question feels familiar, solve it and move on. End the warm-up 10 minutes before the appointment time, because you want a buffer for check-in, lockers, the photo, and the palm-vein scan at the centre — or, for the online format, the room scan, the ID check, and the proctor handshake.
Inside the test: pacing anchors and triage rules
Once the timer starts, the only thing that changes from your last mock is the adrenaline. The pacing budgets are the same: 2 minutes 15 seconds per Quant item, 1 minute 45 seconds per Verbal item, and the family-specific budgets for Data Insights. The first 5 questions of each section are the calibration window, and the adaptive engine is watching both accuracy and timing closely. Do not speed-read the first 5 to bank time. Do not over-solve the first 5 to prove you belong. Sit at the trained pace and let the algorithm do its work.
Around question 8 or 9 of a section, glance at the timer. You should be inside the 30-second buffer you trained with. If you are inside it, continue at the trained pace. If you are outside it on the wrong side, switch from solve-every-question to triage mode for the rest of the section: solve the first 30 seconds, mark and guess if no path appears, and bank the saved time for a single high-confidence question at the end. If you are outside it on the right side, slow down slightly on the harder stems to lift accuracy. Pacing is a feedback loop, and the question-8 or question-9 timer check is the only diagnostic you need inside the section.
Common pitfalls and how to avoid them in exam week
Five pitfalls account for most of the avoidable score damage I see in the final week. The first is the day-six panic drill — the candidate who suddenly decides the night before the test that they have not done enough Two-Part Analysis and grinds through 30 questions at 1 am. The fix is the day-six cap at 60 minutes and the explicit rule that new content stops at day seven. The second is the day-four mock that goes badly and contaminates the next three days. The fix is the conditional mock rule described above: only sit a day-four mock if the previous two mocks were within 10 points. The third is the night-before reorganisation of the error log, where the candidate decides to re-tag every entry by question family and stays up until 2 am. The fix is to lock the error log on day five; it is a record, not a project.
The fourth pitfall is test-morning sugar. The candidate who skips breakfast and drinks two large coffees arrives at the test chair with a sharp glucose and caffeine curve that crashes in the second 45-minute block. The fix is the breakfast protocol above, with protein and slow-release carbohydrate and a single modest coffee. The fifth pitfall is the post-mock Reddit spiral. After a day-four mock, the candidate reads threads about other people's scores and either gets overconfident or gets anxious. The fix is to schedule a non-screen activity for the evening of day four and to stay out of online GMAT communities until after the real test.
Adapting the plan to your test-day format: centre or online
The plan above is largely format-agnostic, but two format-specific adjustments matter. For a centre-based sitting, build in a 15-minute buffer for the check-in queue, the locker, and the palm-vein scan, because arrival-time stress leaks directly into the first five minutes of the test. For an online proctored sitting, do a full dress rehearsal of the room scan, the ID presentation, and the proctor handshake at least 24 hours before the test, because the most common online-format failure is a room-scan issue that forces a re-do and burns 10 to 15 minutes of adrenaline you would rather spend on the first Quant stem.
Online candidates also need to think harder about the optional 10-minute break. The break is the only moment in the sitting when you can stand up, drink water, and reset the pacing rhythm. Use it. The two minutes of stretching, the sips of water, and the 30 seconds of looking out a window are worth more than the two questions you might have banked by skipping the break. Centre candidates have the same break option; the same advice applies. The break is not a luxury, it is part of the section pacing budget.
| Day | Primary block | Time budget | Goal |
|---|---|---|---|
| Seven | Light mixed practice | 90 minutes | Keep pacing hand warm |
| Six | Weak-area drill only | 60 minutes | Sharpen one pattern |
| Five | Full rest or untimed walk-through | 0 to 30 minutes | Let patterns consolidate |
| Four | Mock test (conditional) plus review | 3 to 3.5 hours | Confirm score trajectory |
| Three | Error log and pacing drills | 90 minutes | Build pacing anchor |
| Two | Full rest, logistics only | 0 minutes study | Protect sleep and prep |
| One | Light warm-up, no new content | 15 to 20 minutes | Motor warm-up only |
| Test day | Morning warm-up plus execution | 15 to 20 minutes | Run the trained plan |
What to do if the sitting goes sideways
Even with a clean plan, the sitting can go sideways. The two most common derailments are a first-section pacing slip and a mid-section focus break. For the pacing slip, the move is to accept the slip on the question where it happened and re-anchor at the next question; chasing the lost time across the rest of the section makes the second half worse. For the focus break, the move is to use the next easy question as a reset — a question you can solve in under a minute and a half, with no flags. Solving one easy item cleanly is often enough to bring the rhythm back; a long and tangled item will not.
If the sitting goes badly enough that you are considering a score cancellation on the way out, do not decide in the first 10 minutes after the test. Adrenaline and recency bias distort the picture. Walk for 15 minutes, drink water, eat a small snack, and then make the cancellation decision with a clear head. Most candidates who think they bombed the test have actually scored within 10 points of their last mock; the sitting almost always feels worse than it was. If you do cancel, the score report will not show a numeric score, and the attempt will not count against any school-visible limit, so the cancellation is a free option if used within the window.
Conclusion and next steps
The final 168 hours before a GMAT Focus sitting are a tapering and execution problem, not a learning problem. The plan above — light mixed practice, weak-area drill, conditional mock, error-log review, full rest, logistics check, and a 15-minute test-morning warm-up — is built around the cognitive reality that the brain consolidates during sleep, not during last-minute grinding. Each of the seven days has a single job, and the day's job changes as the test approaches. Most candidates who follow a structure like this arrive at the test chair with the pacing, the focus, and the sleep they need to convert their training into the score they have built toward.
TestPrep İstanbul's GMAT Focus mock-test diagnostic is a natural starting point for candidates who want to confirm where their last full-length adaptive score actually sits before they decide whether day four of exam week should be a mock or a pacing drill.