Blog Test providers compared
SHL vs Aon vs Cappfinity vs Arctic Shores: UK Test Providers Compared
7 min read · by Nicholas Saunders · Published · Last reviewed
Short answer. Four publishers cover most UK finance and consulting graduate assessments, and they are not variations on a theme. Aon publishes its scales numerical test as 37 tasks in 12 minutes with a deduction for every wrong answer. SHL publishes Verify G+ as 30 questions in an allowed 36 minutes, with no deduction. Cappfinity has no clock at all — its assessments are “time recorded, not time limited”. Arctic Shores is not a question paper: it is a sequence of interactive tasks, and it has no overall time limit at all. All four report a percentile rather than a mark, and none publishes a pass score.
This page is the comparison table, with every value labelled by how well it is sourced. Where a publisher states a figure itself we say so and quote it. Where only preparation providers and candidate reports agree, we say that instead. Where nobody publishes it, we write “Not published” rather than fill the gap with a guess — because an estimate written confidently is the thing most likely to be repeated back to you as fact.
How the four providers compare
| SHL | Aon (cut-e) | Cappfinity | Arctic Shores | |
|---|---|---|---|---|
| Publisher | SHL, founded 1977 | Aon Assessment Solutions, formerly cut-e | Cappfinity Limited, founded 2005, traded as CAPP until about 2020 | Arctic Shores |
| Flagship product | Verify range, including Verify G+ General Ability | scales numerical and scales verbal | Immersive and strengths-based assessments, built per employer | Skyrise City task-based assessment, white-labelled as PwC Career Unlocked and KPMG Ready |
| Questions | Confirmed Verify G+: 30 maximum, 10 each of numerical, deductive and inductive. Reported Verify Interactive Numerical: 10. Verify Numerical Ability: 16–18. Verify Verbal: 30. All from preparation providers, none from SHL. | Confirmed scales numerical: 37 tasks. scales verbal: 49 tasks. | Reported Set by the employer. HSBC: 38. KPMG: about 28 across six sections. Deloitte: 14 numerical, 4 verbal, 8 diagrammatic plus situational judgement. | Reported Varies by task: 15 (Order), 20 (Patterns), 45 (Balloon), 50 (Expressions), 80 (Direction). The optional numerical Analysis task is about 18. |
| Time | Confirmed Verify G+: 36 minutes allowed, 32–36 minutes average. Reported Verify Interactive Numerical: 18 minutes. Verify Numerical Ability: about 20 minutes. Verify Verbal: 19 minutes. | Confirmed 12 minutes for the numerical test and 12 for the verbal — roughly 20 and 15 seconds per task. | Reported No limit. Guides suggest about 50 minutes for HSBC’s, about 90 for KPMG’s, and 60–90 for Barclays’ and Deloitte’s. | Not published There is no overall limit and Arctic Shores publishes no duration — the tasks are timed, the sitting is not. Reported Individual tasks: Analysis 18 minutes, Patterns 17. One preparation guide suggests allowing 75–85 minutes for the full suite. |
| Timing model | Reported Strictly timed, with a visible countdown | Confirmed Strictly timed. Aon’s instruction is “finish as many as you can”. | Reported “Time recorded, not time limited” — the on-screen wording, quoted verbatim in candidate walkthroughs. No countdown, but your elapsed time contributes to the score. | Reported Per-task timing, no overall clock |
| Negative marking | Reported None on any Verify test. SHL publishes no marking scheme. | Confirmed Yes. Aon states that “points are usually added for each correct answer and deducted for every incorrect answer”. It does not publish the size of the deduction, and states no penalty for a blank. | Not published None published, and none reported in any candidate walkthrough. | Reported None. Behavioural tasks have no right answer, and the cognitive tasks are scored on performance patterns. |
| Scoring model | Confirmed Percentile against an employer-chosen norm group, plus a sten score on a 10-point scale (mean 5.5, standard deviation 2) | Reported Percentile against an employer-chosen norm group; sten scores in some employer reports | Reported Percentile within the applicant pool for cognitive sections; a strengths profile for the situational sections | Reported A Fit Score percentile against a job-relevant norm group. Cognitive results are banded rather than ranked precisely: below the 30th, 30th to 50th, above the 50th. |
| Answer format | Reported Multiple choice, or interactive — dragging bar heights, setting ranges, ordering items. Verbal is True / False / Cannot Say. | Reported True / False / Cannot Say against data spread across about six tabs. Answering with the wrong tab open scores as incorrect. | Reported Free-text numerical entry, and ranking on an exclusive 1–5 scale with no ties permitted | Reported Tapping, timing and forced choices under pressure. The Analysis task is the only multiple-choice element. |
| Calculator | Reported Permitted on numerical tests, on-screen or physical | Confirmed Not specified in advance. Aon says materials are “generally not required”, that a calculator or pen and paper “is useful for some tests”, and that you are told at the test’s introduction. | Reported Permitted, on-screen or physical, with pen and paper advised | Reported No on-screen calculator. A separate physical calculator is advised for PwC’s numerical section. |
| Adaptive? | Confirmed Verify G+ is adaptive. Reported Standalone Verify tests are largely linear. | Reported Linear | Reported Largely linear, with large item banks | Reported Mostly linear. The Tile Sequence and Balance tasks lengthen as you answer correctly. |
| UK graduate users | Reported Barclays (Verify G+ plus a behavioural assessment) and L.E.K. Consulting. Preparation sources also list most bulge-bracket banks and the Big 4, but those listings conflict with the Cappfinity and Arctic Shores sources below. | Reported Morgan Stanley and Nomura, both on shorter custom builds, per JobTestPrep and AssessmentPass. Aon’s UK graduate footprint is narrower than commonly assumed. | Reported HSBC and KPMG are the two mappings our sources agree on. Barclays, Deloitte and EY are also reported. | Reported PwC, as Career Unlocked. HSBC, NatWest, Citi, Metro Bank and Maven Securities are also reported. |
| Published cut-off score | Not published None | Not published None | Not published None | Not published None |
| Last verified | 10 September 2026 | 10 September 2026 | 10 September 2026 | 10 September 2026 |
How to read this table. Confirmed means the test publisher states it — for Aon, on its candidate preparation page; for SHL, in its Verify G+ Test Fact Sheet; for Cappfinity, in the instruction wording candidates are shown. Reported means the only support is secondary — a preparation site or a candidate walkthrough — and we name it rather than let it read as fact. Not published means nobody publishes it, and we have not filled the gap with an estimate. A secondary figure is not worthless, but it has never been checked by the company that built the test, and you should hold it more loosely than the confirmed rows above.
Which differences actually change how you prepare?
Whether guessing is punished. This is the single largest practical difference, and it points in opposite directions. On SHL, Cappfinity and Arctic Shores, leaving a question blank wins you nothing, so answering everything is sound. On an Aon test, Aon itself states that points are deducted for every incorrect answer, which makes a blind guess on a three-option question a net loss. Practising one habit and then sitting the other test costs you marks before you have read a single question.
Whether there is a clock. Aon runs at about twenty seconds a task and is built so that almost nobody finishes; the skill is triage. Cappfinity has no countdown, which relaxes candidates into spending far too long — and the time is recorded and counts. Those two formats reward almost opposite instincts.
How the answer is entered. SHL’s interactive tests ask you to build the answer — dragging a bar to a height, setting a range — rather than pick from a list. Aon asks you to find the right tab first, and answering with the wrong tab open scores as incorrect even when your judgement was right. Neither mechanic is difficult; both cost time the first few times you meet them.
What the employer sees. All four report a rank, not a mark, so “I got 70% right” tells you nothing on its own. Arctic Shores goes furthest: its cognitive results are banded rather than ranked precisely, and the headline output is a Fit Score against a job-relevant profile. The guide to percentiles and norm groups explains what that ranking actually means.
Which provider will your employer use?
Not published No UK employer publishes which provider it has licensed for a given cycle, and published mappings conflict with each other — several major firms appear under two different providers in different preparation sources, because they use different providers at different stages, or changed between cycles. We have marked the employer row above as reported for that reason, and we would rather say so than tidy the conflict away.
The reliable answer takes thirty seconds: read your invitation email. The sender domain and the test link name the provider. An Aon test is hosted on maptq.com and the invitation mentions Aon or cut-e; an Arctic Shores invitation comes from arcticshores.com and sends you to an app; SHL invitations link to a talentcentral.shl.com address and often include a link to SHL’s own practice tests. If the email names a provider, believe the email over any list, including this one.
Where to start
Whichever publisher you draw, the underlying abilities are the same four: numerical reasoning at pace, verbal inference under the True / False / Cannot Say discipline, abstract pattern work, and situational judgement. What changes is how those abilities are examined — the pacing, the penalties, the mechanics — which is exactly why finding out which one you are sitting matters more than doing more questions.
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 section-by-section breakdown, and an honest starting point.
Common questions
Which test providers do UK graduate employers use?
Four publishers cover most UK finance and consulting graduate assessments: SHL, Aon (formerly cut-e), Cappfinity and Arctic Shores. They differ in length, in whether wrong answers are penalised, and in whether there is a clock at all. Your invitation email names the one you are sitting, and it is the only reliable source: employers change providers by role, by division and by year.
Which UK aptitude test has negative marking?
Aon (cut-e) is the one that publishes it. Aon states that on its performance tests, points are usually added for each correct answer and deducted for every incorrect answer. SHL Verify tests are reported to have no negative marking, Cappfinity publishes none and no candidate walkthrough reports any, and Arctic Shores tasks are not marked right or wrong in that sense. On an Aon test, a blind guess is more likely to hurt you than help you.
Which provider's test is the shortest?
Aon's, by a wide margin. Aon publishes its scales numerical test as 37 tasks in 12 minutes, which is about 20 seconds a task. At the other end, Cappfinity assessments have no clock at all, though your completion time is recorded, and a Cappfinity sitting commonly runs 60 to 90 minutes.
Do any of these providers publish a pass mark?
No. None of the four publishes a cut-off, and no UK employer publishes one either. All four report results as a percentile against a comparison group rather than as a mark, so the same performance can pass at one employer and fail at another. Figures quoted elsewhere come from preparation sites aggregating candidate reports, and should be treated as directional estimates.