How to Improve Your SHL Percentile Score: A 4-Week Study Plan

You cannot raise a percentile you have not measured, so week one begins with a diagnostic under real conditions. Sit one full-length numerical, one verbal and one inductive practice test, timed, in a quiet room, with only the tools the real test allows. Do not pause the clock, do not look anything up, and do not retake the same test later and call it progress. The goal is a baseline you trust.

Record three numbers for each test: questions attempted, questions correct, and questions you skipped or guessed. These three figures tell you very different stories. A candidate who attempts 14 of 20 questions and gets 13 right has a speed problem. A candidate who attempts all 20 and gets 11 right has an accuracy problem. Fixing speed and fixing accuracy call for opposite training, and most people waste weeks drilling the wrong one.

Also note which employer you are preparing for. Different employers commission different combinations of tests, and each sets its own cut-offs, which they rarely publish. Look up the format your target employer uses before you decide how to split your time; our TalentCentral assessment guide shows how candidates usually access the tests.

An error log turns every wrong answer into a specific, fixable skill gap, and it is the single most effective habit in this plan. After each practice set, write down every question you got wrong or guessed correctly, and label the cause. Over a week, patterns appear that no amount of general practice would reveal.

Use a simple spreadsheet or notebook with four columns: question type, what went wrong, the correct method, and a one-line rule for next time. Be specific. “Misread the chart” is not useful. “Used the 2024 column instead of the 2025 column when the question said ‘following year’” is useful, because you can build a habit around it: underline the time reference before touching the data.

Do not log only wrong answers. Questions you got right but took more than twice the average time on are also worth recording, because on a real timed test they cost you the questions at the end. If a type of question regularly eats your time, that is a pacing problem to solve in week three.

The first two weeks should rebuild the skills that every question depends on, not chase speed. Speed built on shaky fundamentals collapses under pressure. For about 45 to 60 minutes a day, work on the core skill for each test type and keep your scratch paper tidy so you can check your own work.

For numerical work, practice the calculations by hand first, then check them with the calculator if one is allowed on your test. Rules on calculators and rough paper vary by employer, so check your invitation instructions rather than assuming. Being able to estimate quickly (is the answer roughly 20 percent of 3,400, so near 700?) lets you eliminate wrong options even before exact calculation.

For verbal work, the biggest score gain usually comes from one habit: answer only from the passage. Candidates with industry knowledge often lose marks by answering from what they know is true in the real world. Our guide to true, false and cannot say questions walks through this in detail, and for the pattern-based test see the inductive reasoning tips.

Week three is where you convert accurate skills into points by controlling the clock. On a timed SHL-style test, the time per question is short, and the test is designed so that many candidates do not finish. Pacing is therefore a skill you can practice, not just a personality trait.

Start by calculating your target pace. Divide the total time by the number of questions, then set a checkpoint every five questions. If a numerical test gives you roughly a minute and a bit per question, a question that has taken you two minutes is no longer a good investment. Make a firm rule: if you have no clear route to the answer after about 20 to 30 seconds of working, mark your best guess, flag it if the platform allows it, and move on.

Run at least three fully timed sets this week, and use sectional drills too: 10 questions in 10 minutes, repeated daily. The sectional work builds the pace; the full tests build stamina. After each, update your error log, noting whether any errors came from rushing. If accuracy falls when speed rises, step the pace back slightly: your percentile depends on correct answers, not on attempts.

The final week is about rehearsal, so every session should look like the real test. Take two or three full-length mock tests at the time of day you are likely to sit the real one, on the device and in the room you will use, with notifications off. This removes surprises and shows whether your pacing rules hold up when the stakes feel real.

Between mocks, do light review rather than heavy new learning. Reread your error log, redo the questions you missed in weeks one to three, and stop new practice 24 hours before the test. Cramming the night before rarely adds points and often adds anxiety.

If nerves are a factor for you, practice a short reset routine in the mocks: one slow breath, shoulders down, then back to the question. Our article on managing SHL test anxiety covers techniques that you can rehearse before test day.

On the day, your job is to execute your pacing and avoid unforced errors, not to learn anything new. The habits below are small, but each one prevents a common way of losing points.

• Read the instructions fully: The examples at the start are not scored in many formats, but they show how questions are laid out. Use the time to settle in.
• Underline the question, not just the data: Many wrong answers come from solving a different question than the one asked. Re-read the final sentence before selecting.
• Write down intermediate steps: On numerical questions, a quick written line saves you from redoing the calculation if you doubt it.
• Use elimination: On multiple-choice items, remove options that are clearly the wrong size, unit or direction before calculating.
• Keep moving after a mistake: Dwelling on a question you think you got wrong costs the next one. Treat each question as independent.
• Do not decode the scoring mid-test: Some formats adapt as you answer, so you cannot reliably tell how you are doing. Focus on the question in front of you.

A plateau usually means you are practicing what you already do well, so change the type of practice, not just the amount. Most candidates see steady gains in the first two weeks, then a flat patch. That is normal and fixable.

First, return to the error log and look for the most frequent category. If most mistakes are misreading, doing more questions will not help; slowing your first read will. If mistakes are method gaps, go back to teaching yourself the method from scratch, then do ten questions on that single skill. If mistakes are mostly late in the test, build stamina with back-to-back sets.

If you want a broader approach to the whole process, from the first online test to the interview, our structured SHL preparation plan shows how to organize your time, and the numerical reasoning guide goes deep on the test most candidates find hardest.