SHL Tests in 2026: What UK Finance and Consulting Applicants Actually Face

If you are applying to graduate schemes at UK banks, consultancies or the Big 4 this cycle, there is a strong chance at least one employer will send you an SHL test. SHL is one of the most widely used assessment providers in UK graduate recruitment, and its tests sit between your application form and any human ever reading it. Most rejections in competitive schemes happen at this stage — before an interview, before a CV review by the hiring team, sometimes within days of submitting.

The problem is that much of what candidates read about SHL tests is out of date. SHL has changed its formats significantly in recent years, and older guides still describe tests that many employers no longer use. This guide covers what the tests actually look like in 2026, how the scoring works, and how to prepare without wasting time.

The SHL test formats you are most likely to meet in 2026

SHL does not run one single test. It licenses a family of assessments, and the employer chooses which version you sit. This is why two friends applying to different firms can both say "I did an SHL test" and describe completely different experiences. The formats you are most likely to encounter:

Verify Interactive Numerical Reasoning. The current flagship numerical format: 10 questions in 18 minutes. Instead of picking A, B, C or D, you build the answer on screen — dragging the height of a bar on a chart, placing values into a ranking, or setting a number range. There are five interactive question types: pie charts, line charts, column charts, number ranges and rankings. Candidates who have only practised multiple-choice questions often find the mechanics themselves cost them time on the first few questions.

Verify Numerical Ability (multiple choice). The more traditional format still used by many employers: typically 18 questions, with time limits that vary by version (roughly 17 to 25 minutes depending on which one the employer has chosen). You are shown a table, chart or graph and asked to calculate a percentage change, compare ratios, work out an average, or extract a specific figure.

Verify G+ General Ability. Increasingly common as a first sift: around 36 minutes covering numerical, inductive and deductive reasoning in one sitting — roughly 30 questions in the multiple-choice version or 24 in the interactive version. The questions are shuffled rather than grouped into sections, so you switch between reasoning types constantly.

Verify Verbal Reasoning. A passage of text followed by statements you must mark as True, False or Cannot Say — strictly on the basis of the passage, not your own knowledge. The discipline of ignoring what you know about the real world is what this test actually measures.

Which version you get is stated in your invitation email. Read it carefully before you practise anything — preparing for the wrong format is the most common self-inflicted error at this stage.

How SHL scoring actually works

Your raw score — the number of questions you answered correctly — is not what the employer sees as your headline result. SHL uses norm-referenced scoring: your performance is compared against a norm group, a reference population chosen by the employer, and reported as a percentile. A result at the 72nd percentile means you scored higher than 72% of that comparison group.

This has two consequences that most candidates never hear about.

First, the same raw score can produce very different percentiles at different firms, because employers choose different norm groups. Being compared against a general graduate population is a lower bar than being compared against candidates for finance roles specifically.

Second, no major employer publishes its cut-off. Preparation providers widely estimate that competitive finance and consulting firms sift around the 80th percentile, with the most selective firms higher still — but these are industry estimates, not published figures. What is certain is that the bar is relative: you are not trying to "pass" in any absolute sense, you are trying to score better than the people applying alongside you.

What the questions actually demand

The mathematics in SHL numerical tests is deliberately unremarkable — percentages, ratios, averages, currency conversions, reading values from charts. Nothing beyond GCSE level. If the maths were the hard part, the test would measure education rather than aptitude.

What the test actually measures is speed and accuracy under pressure. At 10 questions in 18 minutes on Verify Interactive, or 18 questions in around 20 minutes on the multiple-choice versions, you have roughly one to two minutes per question — including reading the data, understanding what is being asked, calculating, and entering the answer. Candidates fail not because they cannot do the arithmetic but because they cannot do it at pace, on unfamiliar data, with a countdown running.

This is also why the tests feel harder than practice materials suggest. Working through questions untimed at your desk is a different activity from answering them against the clock. The gap between those two performances is usually the difference between progressing and being sifted out.

How to prepare without wasting time

Three things reliably help, and none of them are secrets:

Practise the format you will actually sit. If your invitation says Verify Interactive, practising multiple-choice questions builds the wrong reflexes. The interactive mechanics — dragging, ranking, setting ranges — need to feel routine before test day.

Practise timed, always. Untimed practice measures your maths. Timed practice measures what the test measures. From your very first practice session, use the real time limits and hold yourself to them.

Find your real baseline before you drill anything. Most candidates start practising immediately, improve at whatever question type they happen to practise, and never learn where they actually stand relative to the standard employers sift at. A single honest, test-conditions measurement of your current level tells you whether you need two weeks of preparation or two months, and which reasoning type needs the work. Without it, you are training blind.

One thing that does not help: memorising practice questions. SHL draws from large question banks and rotates content, and norm-referenced scoring means you are competing against candidates who are also practising. Familiarity with formats transfers to the real test; memorised answers do not.

Where to start

SHL tests in 2026 reward candidates who know exactly which format they are sitting, understand that the score is relative rather than absolute, and prepare under genuinely timed conditions. All of that starts from one piece of information most applicants never collect: an honest measure of where you currently stand.

Before you start preparing, find out where you actually stand. The Acumen Logic Benchmark is a free, one-shot diagnostic — 20 questions under real timed conditions, taken once, like the real thing. You get a percentile, a per-section breakdown, and a clear picture of what to work on before your first real test.