What Is a Good Cognitive Test Score? Benchmarks by Test and Role

CCAT benchmarks by role

The Criteria Cognitive Aptitude Test has an overall average raw score of 24 out of 50. Criteria Corp publishes role-specific target scores and many employers use their defaults.

Sales development reps and customer success roles typically target 22 and above. Operations and junior analyst roles target 25 and above. Technical product manager roles target 30 and above. Vista Equity Partners portfolio roles, which helped make the CCAT famous on hiring forums, commonly target 33 to 36 depending on seniority. The 90th percentile starts around 38, and anything above 42 is extremely rare.

Wonderlic benchmarks by role

The Wonderlic is 50 questions in 12 minutes and has been in continuous use since 1937. The published population average is 20 out of 50. Targets are grouped by role family.

Unskilled labor targets 10 to 12. Clerical and entry-level office roles target 17 to 21. Skilled trades target 21 to 24. Middle management and most technical roles target 23 to 28. Executive and engineering roles target 27 to 32. The NFL Combine, which made the Wonderlic famous for a generation, averages 24 for quarterbacks historically. Only one player has ever recorded a confirmed 50.

PI Cognitive benchmarks by role

The Predictive Index Cognitive Assessment (PICA) is 50 questions in 12 minutes. Raw scores range from 100 to 450, and the overall average is roughly 250. What makes the PI different is that scores are always interpreted against a Job Target Score set by the employer for that specific role.

Typical entry-level roles target 200 to 225. Experienced individual contributors and team leads target 225 to 275. Senior roles and analytically demanding positions target 275 to 325. Roles that demand complex reasoning under pressure, such as investment roles or senior product management, often target 300 and above. A score 5 to 10 points below target is usually treated as a hard cutoff.

SHL percentile framing

SHL reports results as percentiles rather than raw scores for most of its assessments, including Verify G+ and the Verify Interactive series. This makes benchmarking cleaner because the norm group is built into the number.

Graduate consulting roles typically require the 70th percentile or higher. Professional services firms commonly use the 75th percentile. MBB consulting and bulge bracket investment banking draw the line around the 85th percentile. Anything above the 95th percentile is exceptional and is how candidates from non-target schools break into top-tier interview lists.

Watson-Glaser for legal roles

The Watson-Glaser Critical Thinking Appraisal is the most common assessment for legal training contracts and graduate schemes at Magic Circle and Silver Circle firms. It is 40 questions across five sections in 30 minutes.

The practical cutoff at Magic Circle firms is around 75 percent raw accuracy, which maps to roughly the 80th percentile. Silver Circle firms tend to use similar bars. US law firms tend to use the Watson-Glaser less frequently and the benchmarks vary more. When in doubt, aim for 32 correct out of 40 in practice.

What candidates actually score, from our own attempts

Every benchmark above comes from the test vendors. Vendors have an interest in where they place the bar, and none of them publish what real candidates do the week before an assessment. We can, because we hold 1,473 scored attempts and we re-grade every one of them against the question bank before counting it.

The numbers below are means across the attempts that survived that check. Read them as a practice benchmark, not a population norm. Our candidates are self-selected: they are people who went looking for preparation, which is a different group from everyone who sits a screen cold. If anything, that biases these figures upward.

Two things are worth taking from the table. The first is how tightly the speeded tests cluster. CCAT, Wonderlic and PI Cognitive all land within a mark or two of each other at around 28 out of 50, which is why a raw score means nothing until you know which instrument produced it. The second is how far above that the generous-clock tests sit. Scoring 72 out of 90 on the EIAT is unremarkable. Scoring 36 out of 50 on the CCAT would put you near the top of the distribution.

If you are using these to set a target, add a margin. Practice scores flatter you. You are in a familiar room, you are not being watched, and a failed practice run costs you nothing, which quietly changes how you handle a hard question. Aim to beat your real target by around 15 per cent in practice and treat the gap as insurance rather than ambition.

The score you lose by running out of clock

A target is only useful if you can see what stands between you and it. For most candidates on a speeded test, the answer is not difficulty. It is the end of the paper.

Accuracy on our attempts falls steadily from the first fifth of a test to the last: 84.8 per cent correct on the opening stretch, then 72.1, then 70.7, then 60.4, and 53.3 per cent on the closing fifth. Chance on these tests is 20.7 per cent, so even the worst stretch is well above guessing. Candidates are not collapsing at the end. They are simply reaching fewer questions, and the ones they do reach are answered in a hurry.

That shape has a direct consequence for target-setting. If your practice score sits five marks below where you need it, the instinct is to study harder content. The data says to look at coverage first. Count how many questions you left untouched in your last timed run. If it is more than a handful, your gap is a pacing gap wearing a knowledge gap costume, and it is far cheaper to close.

Why percentile beats raw score

Raw score without a norm group is meaningless. A 28 on the CCAT sounds identical to a 28 on the Wonderlic, but they map to wildly different percentiles because the norm groups are different.

Always ask what percentile your target raw score represents within the specific norm group the employer uses. Some vendors publish norm-group options and some employers pick custom subgroups such as "university graduate applicants" or "early-career technical candidates." The percentile is the only number that survives comparison across tests.

How close to the cutoff is risky

Test-day performance is not your practice average. It is your practice average minus a variance penalty. Nerves, unfamiliar software, and the novelty of the actual test environment usually shave two to four percentile points off your typical practice score.

If your target role requires the 70th percentile, practice until you consistently hit the 85th. That 15-point buffer is the safety margin that absorbs test-day variance. If your practice scores are within 10 percent of the cutoff, you are rolling the dice.

How to read any benchmark you are given, including ours

Three questions will tell you whether a published benchmark is worth anything, and they apply to the vendor numbers above, to the figures a recruiter quotes you, and to our own data.

First, which instrument produced it. A raw score is meaningless without the test name and the question count behind it, and the most common way candidates mislead themselves is comparing a Wonderlic number to a CCAT number as though 50 questions on one is the same as 50 on the other.

Second, which norm group it was measured against. The same raw score sits at very different percentiles depending on whether the comparison is the general working population, graduates, or applicants for that specific role. Applicant norms are the toughest, because the group has already been filtered once. A target quoted without a norm group is an opinion, not a benchmark.

Third, who was measured and why they turned up. Our figures come from people who went looking for practice, which is a motivated group and almost certainly scores above a cold population. Vendor averages come from live assessments and carry the opposite bias in places, since candidates sitting a real screen are under pressure a practice run does not reproduce. Neither is wrong. Both need the caveat attached.

Apply those three questions and most of the score anxiety on hiring forums evaporates. A large share of it is candidates comparing numbers that were never comparable in the first place.

What to do if your practice score is well below target

Focus on format familiarity first, not content. Most of the gap between practice and target for candidates below role targets comes from unfamiliarity with the test structure and unrealistic pacing assumptions. Fix those first.

If the gap is more than 20 percentile points, accept that closing it in a week is unrealistic. Take the test anyway if the application deadline requires it, but plan for a longer prep cycle before your next attempt. For many tests, retakes are restricted to six or twelve months, so the first attempt matters.