KPMG Online Assessment 2026: Numerical, Verbal & SJT Guide
KPMG's online assessment is the make-or-break stage for thousands of UK graduate applicants each year. This guide covers every component — numerical reasoning, verbal reasoning, and the situational judgement test — with format details, scoring, and a structured preparation plan.
KPMG Application Process Overview
KPMG UK is one of the Big Four professional services firms, recruiting approximately 1,000 graduates and school leavers annually across service lines including Audit, Tax, Consulting, Deal Advisory, and Technology & Engineering. Competition is intense — the online assessment is where the majority of candidates are screened out before any human at KPMG reviews your application in depth.
The standard KPMG graduate recruitment journey runs through four stages:
The online assessment (Stage 2) is typically sent within a few days of submitting your application, and you usually have five to seven days to complete it from the invitation date. You complete the tests at home in your own time — but unlike a supervised test, you will be sitting it alongside potentially thousands of other applicants, so the competition is still intense.
KPMG typically sends the online assessment to applicants who meet the basic eligibility criteria (degree classification, UCAS points) without reviewing their CV or motivational answers in detail first. This means your first real filter is the aptitude test — not your grades or experience. A strong CV means nothing if you fall below KPMG's score threshold on the assessment.
KPMG uses psychometric assessments developed by specialist providers. The specific platform may vary by service line and intake year, but the format has consistently included numerical reasoning, verbal reasoning, and a situational judgement element. The total assessment typically takes between 60 and 90 minutes to complete, though some service lines may have longer or shorter versions. Always check the invitation email for the specific format and timing for your application.
Numerical Reasoning Test
The numerical reasoning section is the most demanding component of the KPMG online assessment and the one where most candidates lose ground. It tests your ability to interpret numerical data presented in tables, charts, and graphs — not your ability to do complex arithmetic from memory.
What the Questions Look Like
Each question presents a data exhibit — typically a bar chart, line graph, pie chart, or multi-column table — followed by a multiple-choice question. You must extract the relevant figures, apply a calculation, and select the correct answer from five options. The calculations required include:
- Percentage change: How much has revenue grown from Year 1 to Year 2?
- Ratio and proportion: What is the ratio of Product A sales to Product B sales in Q3?
- Fractions of totals: What fraction of total expenditure is allocated to marketing?
- Rate calculations: If output increased by 12% and workforce fell by 8%, what happened to output per employee?
- Currency conversion and index numbers: Convert figures between currencies or scale an index to a new base year.
| Aspect | Typical Format |
|---|---|
| Number of questions | Approximately 18–20 questions |
| Time allowed | Approximately 25 minutes |
| Time per question | Around 75–85 seconds |
| Question format | Data exhibit + 5-option multiple choice |
| Penalty for wrong answers | No — unanswered questions score zero |
| Calculator | On-screen calculator provided |
| Data exhibits | Tables, bar charts, line graphs, pie charts |
Always read the question first, then scan the data exhibit for the specific figures you need. Most exhibits contain more data than any single question requires. Reading the question first tells you exactly which row, column, or series to focus on — you will save 15–20 seconds per question by not reading the entire exhibit first.
Common Traps to Avoid
KPMG's numerical questions — like most SHL-style tests — are designed with plausible distractors. The wrong answers are not random; they correspond to common calculation errors. The most frequent traps are:
- Percentage change vs. percentage point change: If market share rises from 20% to 25%, the percentage point change is 5, but the percentage change is 25%. Many candidates confuse these.
- Wrong base year: When calculating growth, make sure you are dividing by the starting figure, not the ending figure.
- Unit confusion: Data may be presented in thousands or millions. Missing a scale factor produces an answer that is in the wrong answer list — by design.
- Approximation errors: Rounding intermediate calculations too aggressively can push you to a wrong answer. Use the on-screen calculator for all multi-step problems.
For a deeper dive into numerical reasoning question types and strategies, including worked examples of percentage change and ratio problems, see our dedicated knowledge guide.
Verbal Reasoning Test
The verbal reasoning section tests your ability to understand written information and draw logical conclusions from it — a core skill for any professional services role where you will need to analyse reports, client documents, and regulatory guidance quickly and accurately.
Question Format
You will be presented with a short passage of text — typically 100–200 words on a business, financial, or general topic — followed by a statement. Your task is to decide whether the statement is:
- True: The statement is directly supported by the information in the passage.
- False: The statement directly contradicts information in the passage.
- Cannot Say: You cannot determine whether the statement is true or false from the information in the passage alone — neither supported nor contradicted.
The "Cannot Say" option trips up the majority of test-takers. It does not mean "I'm not sure" — it means the passage provides neither supporting nor contradicting evidence for the statement. If the statement could be true based on your general knowledge but the passage doesn't confirm it, the answer is Cannot Say. You must treat the passage as the only source of truth, ignoring everything you know about the topic from outside the test.
| Aspect | Typical Format |
|---|---|
| Number of questions | Approximately 18–20 questions |
| Time allowed | Approximately 17–19 minutes |
| Time per question | Around 55–65 seconds |
| Passages per set | Typically 1 passage with 3–4 questions each |
| Answer options | True / False / Cannot Say |
| Topics covered | Business, economics, general professional content |
Strategies That Actually Work
Unlike the numerical test, the verbal reasoning test rewards a disciplined mental approach more than any specific knowledge. The following habits separate high scorers from average scorers:
- Locate the relevant sentence before answering: After reading the statement, go back to the passage and find the sentence or clause that is most directly relevant. Do not answer from memory — reread that sentence carefully before choosing.
- Flag absolute language: Words like "always", "never", "all", "none", "the only" in a statement are almost always False or Cannot Say, because the passage rarely contains equally absolute support for them.
- Separate inference from statement: If the statement is a logical inference that could follow from the passage but is not explicitly stated, the answer is Cannot Say — not True. The test is about what the passage says, not what it implies.
- Keep pace — do not linger: At roughly one minute per question, you cannot afford to spend three minutes on a difficult item. Make your best assessment, flag it mentally, and move on. Coming back rarely helps and costs time you do not have.
For detailed worked examples of the True/False/Cannot Say format, including the most common types of tricky statements, see our guide: Verbal Reasoning: Mastering True, False and Cannot Say.
Situational Judgement Test (SJT)
The situational judgement test is the component most candidates underestimate — and the one that most clearly reflects KPMG's values and culture. Rather than testing cognitive ability, the SJT presents you with realistic workplace scenarios and asks how you would respond. There are no objectively correct answers in the way a maths question has one, but there are better and worse responses in relation to KPMG's competency framework.
What a Typical SJT Scenario Looks Like
Each scenario describes a situation you might plausibly encounter as a new KPMG graduate — a challenging client interaction, a team conflict, an ethical dilemma, a difficult deadline, or a situation where you have received confusing instructions from a manager. You are then asked to either:
- Rate each possible response: On a scale from "Highly Effective" to "Counterproductive" — you evaluate multiple response options independently.
- Rank the responses: Put the response options in order from most to least appropriate, which forces you to make relative judgements.
KPMG's SJT is designed around their publicly stated values: Integrity, Excellence, Courage, Together, and For Better. The "correct" responses in each scenario are those that reflect these values in a professional context — particularly integrity in ethical situations, collaboration in team scenarios, and a client-first mindset in service delivery situations. Understanding KPMG's values before you sit the SJT gives you a meaningful interpretive framework.
How to Approach Each Scenario
The SJT is not about choosing the most dramatic or impressive action — it is about demonstrating sound professional judgement. The following principles will consistently guide you toward the higher-scoring responses:
- Prioritise communication and transparency: In almost every scenario, responses that involve communicating with the relevant people (your manager, the client, your team member) score higher than those that involve acting unilaterally or ignoring the issue.
- Ethical situations: always flag upward: If a scenario involves anything that could be an ethical breach — a colleague cutting corners, a client asking for something inappropriate, a discrepancy in figures — the correct response is always to escalate to a more senior person, not to handle it yourself or stay silent.
- Avoid extreme responses: Responses that are either completely passive ("do nothing, hope it resolves itself") or aggressively confrontational ("immediately report the colleague to HR") are almost never the highest-scoring option. Measured, professional, constructive responses score best.
- Consider the client and the firm: KPMG is a client services firm. Responses that protect client relationships and the firm's reputation alongside individual integrity will typically score above those that consider only one stakeholder.
| Scenario Theme | What It Tests | Key Principle |
|---|---|---|
| Deadline pressure with quality risk | Integrity vs. expediency | Never compromise quality; escalate early |
| Colleague acting unethically | Courage and integrity | Address it directly and/or escalate |
| Client dissatisfied with work | Client focus and professionalism | Listen, acknowledge, resolve — don't deflect |
| Team conflict or disagreement | Collaboration and communication | Seek to understand, find common ground |
| Ambiguous or unclear task | Initiative and communication | Clarify before acting, not after |
| Workload conflict or prioritisation | Time management and communication | Communicate constraints proactively |
For a comprehensive overview of the situational judgement test format — including the difference between rating and ranking formats — see our SJT knowledge guide.
How KPMG Scores Your Results
KPMG does not publish its specific scoring methodology or cut-off thresholds. What is publicly known — from KPMG's own recruitment communications and candidate feedback — is the following:
- Each test component is scored separately. Your numerical, verbal, and SJT scores are evaluated independently. A strong SJT score does not compensate for a very low numerical score — you need to perform adequately across all three components.
- Scores are standardised against a norm group. Like all major psychometric assessments, your raw score is converted into a percentile rank or standardised score that reflects your performance relative to a relevant comparison population — typically graduate-level candidates.
- Cut-off scores are not shared publicly. KPMG does not publish the specific threshold scores required to progress. Based on candidate reports and the competitive level of KPMG's intake, you should aim to perform above the 65th percentile on each component as a minimum target, and ideally above the 70th–75th percentile to give yourself a comfortable margin.
You will find websites and forums claiming that KPMG requires a specific score (e.g. "70% correct" or "sten 7"). These figures are not publicly confirmed by KPMG and may be outdated, inaccurate, or referring to a different service line or assessment version. Cut-off thresholds also change year to year based on applicant volume and intake targets. Focus on maximising your performance rather than targeting a specific number.
The SJT Scoring Mechanism
For the situational judgement test specifically, scoring works differently from the cognitive tests. Rather than a simple right/wrong count, SJT responses are typically scored against an "expert key" — the responses that a panel of experienced KPMG professionals agreed represented the most and least effective professional behaviour. The closer your ratings or rankings are to the expert key, the higher your SJT score.
Because the SJT is designed to assess values alignment and professional judgement — not numerical ability — there is less variance between candidates on this component than on the cognitive tests. The cognitive tests (numerical and verbal) tend to be the primary differentiators in the screening process.
Does the Test Differ by Service Line?
The core format — numerical reasoning, verbal reasoning, and SJT — is broadly consistent across KPMG's graduate service lines. However, there are some differences in emphasis, and it is worth being aware of how the role you are applying for may shape what the assessment is testing for.
| Service Line | Primary Skills Assessed | Relative Emphasis |
|---|---|---|
| Audit | Numerical accuracy, attention to detail, professional integrity | High numerical + SJT weight |
| Tax | Verbal reasoning, legislative interpretation, logical thinking | High verbal + numerical balance |
| Consulting | Structured problem solving, client communication, data interpretation | Balanced across all three components |
| Deal Advisory | Financial analysis, numerical precision, commercial judgement | Strong numerical emphasis |
| Technology & Engineering | Logical reasoning, technical aptitude, problem solving | Balanced; some roles may include additional logical/diagrammatic elements |
Read KPMG's service line pages and the job description for the role you are applying to. The language KPMG uses — "precision and accuracy" for Audit, "commercial insight" for Consulting, "analytical rigour" for Deal Advisory — tells you which competencies they are emphasising. This is directly relevant to how you approach the SJT scenarios, where understanding the role context helps you identify the most professionally appropriate response.
KPMG's school leaver and apprenticeship programmes use the same core assessment format as the graduate scheme but may be calibrated to a different norm group. If you are applying as a school leaver, the cognitive content is the same type — but you are being compared against a school leaver population rather than a graduate population, which changes the percentile implications of your raw score.
4-Week Preparation Plan
Four weeks of structured preparation is enough to produce meaningful improvement on all three components of the KPMG online assessment — provided the practice is deliberate, timed, and followed by honest review. Here is a week-by-week plan:
Complete one full timed numerical, verbal, and SJT practice test under exam conditions (no pausing, no looking things up). Record your scores. This is your baseline. Then spend time reviewing every question you got wrong — not just the correct answer, but exactly why your reasoning led you to the wrong one. In the second half of the week, study the core formula toolkit for numerical reasoning: percentage change, ratios, fractions of totals, and rate calculations. Drill these until they are automatic.
Focus the bulk of this week's practice on numerical reasoning — typically the highest-variance component. Complete a minimum of three timed numerical tests, reviewing each one in full. Work specifically on the question types you got wrong in Week 1. Pay particular attention to percentage change questions (the single most common trap) and multi-step calculations involving currency conversion or index numbers. By the end of Week 2, you should be answering most straightforward numerical questions within 60 seconds.
Shift emphasis to verbal reasoning. Complete two to three timed verbal tests, concentrating specifically on Cannot Say questions — these are where most marks are lost. After each test, analyse every Cannot Say question: was it the passage's silence that made it Cannot Say, or did you miss a clause that actually made it True or False? For the SJT, read through KPMG's values on their website (Integrity, Excellence, Courage, Together, For Better) and practise mapping each scenario to the relevant value before selecting your response.
Simulate the full KPMG online assessment in one sitting: numerical + verbal + SJT back to back with no break. This builds the mental stamina the real test requires. Do this twice in Week 4. Between simulations, focus only on your remaining weak spots — if verbal is still lagging, drill that; if percentage change questions are still catching you, drill those. In the final two days before your real test, do lighter revision only — one or two practice tests at most. Do not cram the night before. Rest, sleep well, and approach the assessment fresh.
Untimed practice builds accuracy but not the speed and pressure management the KPMG assessment requires. From your very first practice session, set a timer and replicate the actual time limits. Test anxiety and time pressure are the biggest performance gaps between practice and real test scores — the only way to close them is to practise with the pressure on. Candidates who practise only in relaxed, untimed conditions are often shocked by how differently they perform when the clock is running.
For a broader overview of what to expect across the full graduate assessment process — including the virtual assessment centre — see our Big Four assessment centre guide, which covers the group exercise, written exercise, and interview formats common to KPMG and the other Big Four firms.
You can also benchmark your current level with our free SHL-format practice tests, which use the same numerical and verbal reasoning question types as the KPMG online assessment.
Frequently Asked Questions
Prepare for the KPMG Online Assessment
Access our full library of SHL-format numerical, verbal, and situational judgement practice tests — the same question types used in the KPMG online assessment, with detailed explanations and score analytics.