How to prepare for the Revolut operations analytical test

Landing an operations role at Revolut means surviving one of fintech's most notoriously rigorous screening steps. Long before you speak to a hiring manager or present a case study, you will face the operations analytical test. Revolut operates on razor-thin margins and massive global scale. They need operators who can process complex data, calculate probabilities, and apply structured logic—all under extreme time pressure. The company designed this assessment specifically to filter out candidates who struggle with rapid, data-backed decision-making. Passing takes a lot more than basic math skills. You need a ruthless approach to time management, a rock-solid grasp of analytical frameworks, and the ability to estimate accurately when the clock is ticking. This guide breaks down exactly what you will face, how to shortcut complex math, and the specific frameworks that will get you through the hardest questions.

The Revolut operations hiring process

Before diving into the test mechanics, you need to understand where this assessment sits in the broader hiring funnel. The analytical test is the ultimate bottleneck. Revolut recruiters use it to aggressively thin the herd before committing human hours to case interviews. If you fail here, your application ends immediately, regardless of your resume or pedigree. Knowing the stakes helps you prioritize your preparation time effectively.

Format and structure of the assessment

You will take this test through an internal portal or a third-party platform like SHL or HackerRank. Expect a tight timeline. You will face 15 to 20 questions and only have 20 to 30 minutes to finish them. The format leans heavily on multiple-choice questions, but you will likely have to type exact numerical answers for a few free-text fields. The assessment evaluates three core pillars: numerical reasoning, probability and statistics, and logical deduction.

What really defines this test is the intense time pressure. You have roughly 60 to 90 seconds per question. Hesitation is not an option. Furthermore, the testing platform usually prevents you from going back to previous questions. Once you submit an answer, it is locked. This means you have to constantly balance accuracy with forward momentum. You cannot afford to get stuck on a single problem for four minutes, because doing so mathematically guarantees you will not finish the test.

How do I solve Revolut probability questions?

Probability questions here mimic real-world risk, fraud, and user behavior scenarios. Let's look at a classic false-positive example: 'If 4% of all transactions are fraudulent, and our internal model correctly flags 95% of fraudulent transactions but has a 2% false positive rate, what is the probability that a flagged transaction is actually fraud?'

If you want to solve this quickly, skip Bayes' theorem formulas. Use a hypothetical base number instead, like 10,000 transactions. Out of that 10,000, 4% means 400 are fraud and 9,600 are legitimate. The model catches 95% of those 400, meaning 380 are flagged as fraud. It also falsely flags 2% of the 9,600 legitimate ones, giving you 192 false flags. Add those together (380 + 192) and your total flagged count is 572. The probability the transaction is actually fraud is 380 divided by 572, which comes out to roughly 66.4%. Sticking to concrete numbers prevents clumsy algebraic mistakes when the clock is ticking.

Let's try a second scenario involving compound probability. 'A user has a 20% chance of referring a friend. If they refer a friend, that friend has a 30% chance of opening an account. If they open an account, there is a 50% chance they fund it within 7 days. What is the probability that a single existing user generates a funded account within 7 days?'

Multiply the probabilities: 0.20 * 0.30 * 0.50. First, 0.20 * 0.30 = 0.06. Then, 0.06 * 0.50 = 0.03. The answer is 3%. When dealing with compound events, convert percentages to decimals or fractions immediately on your scratchpad. 1/5 * 3/10 * 1/2 = 3/100, which is instantly recognizable as 3%.

Finally, you must understand Expected Value (EV). You might see a prompt like: 'You are launching a new feature. There is a 20% chance it increases revenue by €100,000, a 50% chance it does nothing (€0), and a 30% chance it causes a bug that costs €40,000 to fix. What is the expected value of the launch?' Calculate the weighted average of all outcomes: (0.20 * 100,000) + (0.50 * 0) + (0.30 * -40,000). This becomes 20,000 + 0 - 12,000, resulting in an expected value of €8,000. Expected value questions are guaranteed to show up because they form the baseline of operational risk decisions.

How do I answer data interpretation questions?

These questions throw dense tables or multi-axis charts your way. They usually show metrics like user growth, customer acquisition cost (CAC), lifetime value (LTV), or revenue across various European markets. You might see a prompt asking, 'Which region achieved the highest return on investment (ROI) in Q3?'

The trick here is knowing when to estimate and when to calculate. Imagine a table with three countries. The UK has a CAC of £15 and an LTV of £60. France has a CAC of £20 and an LTV of £90. Germany has a CAC of £10 and an LTV of £35. Instead of calculating the exact ROI percentage for every row, look at the multiples. UK is 4x (60/15). France is 4.5x (90/20). Germany is 3.5x (35/10). France wins. You can do this visually without touching your calculator.

Another common data interpretation format involves reading currency conversions mixed with percentage growth. For example, 'Market A generated €4 million in Q1, growing at 15% quarter-over-quarter. Market B generated £3.5 million in Q1, growing at 10%. If the exchange rate is £1 = €1.15, which market will generate more revenue in Q2 in Euros?'

Notice how breaking the problem into isolated, sequential steps prevents you from mixing up growth rates and exchange rates. Write each step down clearly on your scratchpad to avoid losing your place.

Logical reasoning: Deciphering rules under pressure

Revolut operations managers constantly map out product flows, edge cases, and compliance rules. To test your aptitude for this, the assessment includes strict logical deduction questions. You will get a set of absolute rules and have to determine which outcome is definitively true based only on the provided text.

Example: 'To issue a metal card, a user must be on the Premium plan for at least 3 months OR have a monthly transaction volume above €5,000. Users with a frozen account cannot receive any physical cards. John has a monthly volume of €6,000, has been on the Standard plan for 1 year, and his account is active. Can John receive a metal card?'

Break down the logic gates. Condition 1: Premium for 3 months (False) OR Volume > €5,000 (True). Since it is an OR statement, the first gate is passed. Condition 2: Account not frozen (True). Therefore, John can receive the card. The difficulty ramps up when they introduce conditional statements with negative phrasing ('Unless X happens, Y will not occur'). When you encounter these, draw a quick flow chart on your scratchpad. Map the 'If/Then' relationships visually. Do not try to hold multiple overlapping rules in your working memory.

How do I tackle market sizing and estimation?

You will definitely see estimation logic in the live case interviews, but it also pops up frequently in the analytical test. A typical question asks, 'Estimate the daily transaction volume of Revolut in the UK.' This is where you apply the MECE (Mutually Exclusive, Collectively Exhaustive) framework. Break the problem down step-by-step using round numbers.

Because you are dealing with a multiple-choice format, a rough order-of-magnitude calculation is usually enough. The options might be 200,000, 2 million, 20 million, and 200 million. By simply structuring the logic and using friendly numbers (like 70 million instead of 67.3 million), you land squarely on the correct magnitude without doing any precise arithmetic.

How to set up your testing environment for maximum speed

Speed is entirely dependent on your physical setup. Do not take this test on a 13-inch laptop screen while sitting on your couch. You need an optimized environment designed to shave seconds off every interaction.

Common traps and prep strategies

Poor time management is the number one reason candidates fail this assessment. Revolut intentionally designs the test so most people will not finish every question. If you waste four minutes staring at a complex probability tree, you just sacrificed three much easier data interpretation questions waiting at the end.

Trap 1: The Perfectionist Calculator. Candidates often calculate answers to the third decimal point. The multiple-choice options are usually spaced far enough apart that rough estimation gets you the right answer in half the time. If you need to multiply 4,192 by 18.7%, just multiply 4,200 by 20% (840) and pick the answer slightly below that.

Trap 2: Misreading the units. Data tables love to hide unit variations. The revenue column might be in thousands (000s), while the user column is in absolute numbers. If you divide revenue by users without adding the zeros back in, your ARPU (Average Revenue Per User) calculation will be off by a factor of a thousand. Always read the axis labels and table headers before looking at the raw data.

Creating a 7-day study plan for the Revolut test

Cramming the night before will not work for a skills-based test. You need to build muscle memory over a week so that your brain naturally defaults to the correct frameworks under pressure.

Transitioning to the live rounds with AcePrompt

Passing the analytical test is a huge milestone, but the real challenge begins right after. You will have to defend your logic in live case interviews and behavioral rounds, which is where communication and structured storytelling become absolutely essential. AcePrompt acts as your real-time AI interview copilot during these live sessions. It actually listens to the interviewer's questions and instantly provides structured, personalized frameworks directly on your screen. Whether you need a STAR format for behavioral questions or a MECE breakdown for a case, the guidance is right there. It ensures you never lose your train of thought, helping you deliver the concise, highly structured answers that operations hiring managers expect.