What the Bain TestGorilla test is
TestGorilla is a hiring platform with a library of short skills tests. An employer picks a handful, and candidates take them in one session. Bain's careers page says its digital assessment runs on either TestGorilla or SOVA, that the email after you apply says which, and that it takes about 30 to 40 minutes. Bain's Seoul recruiting FAQ, which covers the TestGorilla version, gives about 52 minutes, likely including setup and instructions.
Candidate reports place TestGorilla mainly in US offices, parts of Europe (Paris and Amsterdam are named) and Seoul. If your email says SOVA instead, the Bain SOVA guide is the one you want.
The two reported mixes
Bain doesn't publish which TestGorilla tests it uses. Candidates and prep sites report two combinations. The timers and question counts come from TestGorilla's own pages for each test, where TestGorilla publishes them; counts marked "reported" come from prep sites.
| Mix | Tests | Test time |
|---|---|---|
| Europe (reported for Paris and nearby offices) | Numerical reasoning (10 min, 12 q) · Critical thinking (12 min, about 16 q reported) · Working with data (10 min, 12 q) | 32 min |
| US | Numerical reasoning (10 min, 12 q) · Business judgment (10 min, about 15 q reported) · Leadership and people management (10 min, about 12 q reported) · Problem solving (9 min, 12 q) | 39 min |
Some US candidates report three of the four rather than all four, with numerical always included. That comes from one cluster of prep sites that share wording, so treat it as a single report. The order of the tests isn't published anywhere we could find.
What this means for pacing: most tests give you 40 to 50 seconds a question. That's enough time to do the arithmetic carefully once. It's not enough time to do it twice.
The rules that shape your strategy
- Each test has its own timer. Finishing numerical early doesn't give you more time on the next test.
- No going back. Bain's Seoul FAQ and TestGorilla's candidate help both say you can't return to a question once you've moved on. A couple of prep sites say otherwise; the platform's own help is the better source.
- No penalty for wrong answers. TestGorilla's candidate help says wrong answers don't cost you anything extra. Never leave a question blank. If the timer is about to run out, pick your best guess.
- Calculator, pen and paper are allowed. Bain's page says so. Use a physical calculator you know well rather than switching windows.
- One sitting, webcam on. The Seoul FAQ describes webcam monitoring. Sit somewhere quiet, alone, and close everything else.
Numerical reasoning: worked example
Twelve questions in ten minutes, in both mixes. Expect tables and charts with business figures and questions about percentage change, ratios, shares and simple forecasts. The math is school-level. The difficulty is reading the right number from the right row quickly.
Example: fastest growth
A company's revenue by region, in $ millions:
| Region | 2024 | 2025 |
|---|---|---|
| North | 4.0 | 4.8 |
| South | 6.0 | 6.9 |
| West | 2.5 | 3.2 |
Which region's revenue grew fastest from 2024 to 2025?
North grew 0.8 on 4.0, which is 20%. South grew 0.9 on 6.0, which is 15%. West grew 0.7 on 2.5, which is 28%. Answer: West.
The trap: South added the most dollars, so a quick glance picks South. "Grew fastest" means percentage growth unless the question says otherwise. When a question says "fastest", "most", or "largest increase", pause for one second and ask whether it means absolute or relative.
Example: percent versus percentage points
A brand's market share rose from 12% to 15%. By what percentage did its market share increase?
The share rose by 3 percentage points. As a percentage of the starting share, that's 3 ÷ 12 = 25%. Answer: 25%. The option "3%" will be there, because it's the mistake people make.
Speed habits that help: write the two numbers you need before touching the calculator, round only at the end, and check that your answer is the right size (a 25% rise on a 12% share lands at 15%, which matches).
Critical thinking: worked example
Reported for the Europe mix: twelve minutes, about sixteen questions. Expect short arguments and statements where you decide what follows, what's assumed, or what would weaken a claim. Some questions are pure logic.
Example: what must be true
"Every project with a budget over $2 million needs a partner's sign-off. Project Lark needed a partner's sign-off."
Which of these must be true?
- Project Lark's budget was over $2 million.
- Project Lark's budget was $2 million or less.
- Neither A nor B can be concluded from the statements.
Answer: C. The rule says big projects need sign-off. It doesn't say only big projects need it. Lark could have needed sign-off for another reason. Choosing A is the classic error of running a rule backwards.
The habit: before looking at the options, restate the rule as "if X then Y" and remember that Y happening tells you nothing certain about X.
Working with data: worked example
Reported for the Europe mix, and TestGorilla lists it at twelve questions in ten minutes. It's less about calculation and more about filtering: find the rows that meet several conditions, then count, sum or compare.
Example: count with conditions
| Order | Region | Value | Shipped |
|---|---|---|---|
| 101 | East | $620 | On time |
| 102 | West | $540 | Late |
| 103 | East | $480 | Late |
| 104 | East | $910 | Late |
| 105 | North | $700 | Late |
| 106 | East | $505 | Late |
How many East orders worth more than $500 shipped late?
East orders: 101, 103, 104, 106. Of those, late: 103, 104, 106. Of those, over $500: 104 and 106. Answer: 2.
Apply the most restrictive condition first when you can, and write down the survivors after each filter. It's slower than eyeballing and faster than redoing it.
Problem solving: worked example
Reported for the US mix. TestGorilla lists it at twelve questions in nine minutes, the tightest pace of any test here. Expect scheduling puzzles, rules to apply, and small logic problems.
Example: a schedule
Four meetings, A, B, C and D, fill the 9, 10, 11 and 12 o'clock slots. A is before C. B isn't at 9. D is immediately after B. C isn't at 12.
Which meeting is at 10?
B can't be at 9, and it can't be at 12 because D has to follow it. So B is at 10 or 11.
If B is at 10, D is at 11, and A and C take 9 and 12. A must be before C, so C is at 12, which breaks the last rule. So B is at 11 and D at 12. A and C take 9 and 10, with A first. Answer: C.
The method is to place the most constrained item first (here B, because it drags D with it) and test the two cases.
Business judgment: worked example
Reported for the US mix: ten minutes, about fifteen questions. Expect short business scenarios where you pick the most sensible decision or the most relevant piece of information. Some involve light numbers.
Example: a price cut
A product sells 10,000 units a month at $50. Each unit costs $30 to make and ship. The sales team proposes a 10% price cut and expects volume to rise by 20%. What's the best response?
- Approve it: revenue goes up.
- Reject it: monthly profit falls.
- Approve it: more customers is always worth it.
- There's not enough information to decide anything.
Today: 10,000 × ($50 − $30) = $200,000 a month. After: 12,000 × ($45 − $30) = $180,000. Revenue does go up, from $500,000 to $540,000, which makes A tempting. Answer: B. Profit falls, because the cut takes a quarter of the margin on every unit and the extra volume doesn't make up for it.
Judgment questions reward looking at what the business cares about (usually profit) rather than the number that moved most.
Leadership and people management: worked example
Reported for the US mix: ten minutes, about twelve questions. This is a situational judgement test. You read a workplace situation and choose the best response. Several options will sound reasonable, and some questions are genuinely judgement calls.
Example: missed deadlines
You lead a team of four. One analyst, usually reliable, has missed two deadlines in two weeks. The client hasn't noticed yet. What do you do first?
- Reassign their work to others to protect the client.
- Talk to them privately to understand what's changed.
- Raise it with your manager so there's a record.
- Mention it in the team meeting as a reminder about deadlines.
Best answer: B. A usually reliable person missing deadlines is a change, and you don't know the cause yet. It might be workload, a personal issue, or unclear instructions from you. A private conversation finds out and keeps the person's trust. Reassigning (A) might be right later, but doing it first treats a symptom. Escalating (C) before talking to them skips a step. Calling it out in a group (D) embarrasses one person to send a message to everyone.
The pattern in these tests: prefer responses that are direct, fair, and deal with the cause, and be wary of options that avoid the conversation or escalate before you know what's going on.
How it's scored
Bain doesn't publish a pass mark, and its Seoul FAQ says results aren't shared with candidates. On the platform side, TestGorilla reports each candidate's result as a percentile against other people who took the same test, adjusted for question difficulty. It switched to this in October 2025, so older guides that describe scores of "1 to 3 per sub-skill" are describing a system that's gone.
What that means for you: there's no fixed number of correct answers that passes. You're being compared with other candidates, so every extra correct answer helps, and a guess costs nothing.
Common mistakes
- Absolute versus relative. Picking the biggest dollar change when the question asks for the fastest growth, and the reverse.
- Percent versus percentage points. A rise from 12% to 15% is 3 points and 25%.
- Wrong base. Percentage change is divided by the starting value, not the ending one.
- Running a rule backwards. "If big, then sign-off" doesn't mean "if sign-off, then big".
- Spending two minutes on one question. With no going back, a stuck question should get your best guess and you should move on.
- Leaving blanks. There's no penalty for a wrong answer, so a blank is always worse than a guess.
- Preparing for the wrong mix. If you're in Europe, critical thinking and data questions matter more; in the US, judgement and problem solving do. Numerical is in both.
A one-week plan
- Day 1: confirm TestGorilla in your email and note your office. Do one untimed practice set of each test in your likely mix to see where you're slow.
- Day 2: numerical. Percentage change, points versus percent, ratios, reading two-row tables. Use the calculator you'll use on the day.
- Day 3: numerical again, now at 50 seconds a question. Don't check answers mid-set.
- Day 4: critical thinking and working with data (Europe) or problem solving (US). Practise the "restate the rule" and "most constrained first" habits.
- Day 5: business judgment and people management (US), or more data filtering (Europe). For judgement questions, read the explanations and learn why an answer is stronger, not only which one.
- Day 6: one full mock, tests back to back, timed, no going back.
- Day 7: rest. Check your webcam, calculator and a quiet room. Sit it when you're fresh.
Bain practice on SolveForge has timed numerical, logical, verbal and situational judgement sections, and a TestGorilla-style full mock for pass holders. The questions are ours, written in the reported style, not TestGorilla's.
FAQ
Which tests are in the Bain TestGorilla assessment?
Bain doesn't publish the list. Candidates report two mixes: in Europe, numerical reasoning, critical thinking and working with data; in the US, numerical reasoning, business judgment, leadership and people management, and problem solving. Numerical reasoning is reported in both.
How long is the Bain TestGorilla test?
The reported mixes add up to about 32 minutes (Europe) or 39 minutes (US) of test time. Bain's careers page says about 30 to 40 minutes; its Seoul FAQ says about 52 minutes, likely including setup.
Can you go back on Bain's TestGorilla test?
No. Bain's Seoul FAQ and TestGorilla's candidate help both say you can't return to a question after moving on. Each test also has its own timer.
Are wrong answers penalised on TestGorilla?
No. TestGorilla's candidate help says wrong answers aren't penalised, so answer every question, even with a guess.
Can I use a calculator on the Bain TestGorilla test?
Yes. Bain's careers page says pen, paper and a calculator are allowed.
What is a good score on the Bain TestGorilla test?
Bain doesn't publish a pass mark and doesn't share results with candidates. TestGorilla reports results to employers as percentiles against other test takers, so the aim is simply as many correct answers as you can get.
Keep going
Back to the Bain online assessment overview. The other Bain platform: the Bain SOVA guide. More worked examples: numerical, logical and situational judgement. Applying elsewhere too: the MBB online assessments compared.
Sources we checked: Bain's digital assessment careers page, a Bain Seoul recruiting FAQ, TestGorilla's pages for each test and its candidate help and percentile-scoring articles, and prep-site reports from Prepmatter, Hacking the Case Interview, IGotAnOffer and a PrepLounge candidate post from June 2026.
SolveForge is an independent practice tool and is not affiliated with, endorsed by, or connected to Bain & Company or TestGorilla. Every example on this page was written by us; none is a real test item. "Bain" and "TestGorilla" are trademarks of their owners. Practising here does not guarantee any assessment result.