What you are actually buying

Shown, not described. Every screenshot on this page is the real product as it looks today, every number is checkable, and the sample drill at the bottom is the genuine article — free, no account.

25

teaching drills across Numerical, Verbal, Logic and Situational Judgement

87

named failure modes — the specific errors the drills diagnose and train out

16

replica assessment forms built to published provider specifications — SHL, Aon, Cappfinity

100%

of explanations machine-checked before every deploy — the build fails if an arithmetic claim is wrong

No invented statistics, no "students improve by X%" claims. If a number is on this page, you can count it in the product.

The teaching loop

We diagnose the mistake, not just mark the answer

Most practice sites tell you "incorrect" and show the right answer. That trains nothing. Here is what happens instead, with a real question from the numerical track:

1 · You answer

"Headcount rose from 40 to 52. What was the percentage increase?" You pick 23.1% — it looks right, you divided the change by a real number from the question.

2 · Your specific error is diagnosed

Not "incorrect" — this: "23.1% divides the change by the NEW value (12 ÷ 52). Percentage change always divides by the original — 40." The option you chose was written to catch exactly that move.

3 · The trap is named

The wrong base — 1 of 87 named failure modes. Once an error has a name, you can see it coming in a question you have never met before.

4 · The fix is drilled

The analytics track which traps keep catching you and point you at the drill that trains each one — through worked examples, hinted practice, and a beat-the-clock unit at test pace.

Drill screen: a weighted-average question answered correctly, followed by a numbered step-by-step method, a key takeaway and a think-faster tip.

A drill explanation as it appears in the product: the method step by step, the takeaway, and a think-faster tip.

The trap taxonomy

Your errors, tracked by name

Every wrong option across the drills is tagged with the failure mode it was written to catch — dividing by the wrong base, confusing per cent with percentage points, applying a reverse percentage forwards. The analytics count which ones catch you, show which you have mastered, and link each recurring trap to the drill that trains it.

This is the difference between practising and improving: a score tells you that you lost marks; a named trap tells you why, and what to do about it.

Traps analytics panel: traps mastered with three correct answers in a row, alongside recurring traps — the named error patterns that have caught you most often, each linked to its drill.

The full loop

Benchmark → diagnose → train → track

It starts free: the 15-minute Benchmark scores you against the field and breaks your performance down by category. Pro then adds the replica assessments, the drills, and the analytics that connect them.

Benchmark results screen: score, estimated percentile and time taken side by side, a note on how employers rank applicants against a norm group, and a section breakdown across logical, numerical and verbal reasoning with pace per question measured against a target.

The free Benchmark report — score, estimated percentile, and a section breakdown with pace per question.

Dashboard after practice: standing at the 87th estimated percentile, up 38 since the Benchmark, with the latest sitting an SHL Verify Numerical Reasoning replica scored at 83% and four practice sessions logged this week.

The dashboard as practice accumulates: estimated standing, latest replica sitting, and the week's work.

Don't take the page's word for it

Try a complete drill, free.

Percentages & Percentage Change — the first numerical drill in the programme, with the same lesson and the same named-trap diagnosis Pro members get. No account, no email.

Try the sample drill Start Free Benchmark

Pro is a one-time payment — £79 for 90 days at the founding price (list £147) until 31 October 2026. No subscription, nothing auto-renews. Full pricing