Kiwi Payments Data Analyst Interview: Rounds, Questions and Preparation

What the job involves

Data Analysts at Indian unicorns and high-growth startups like Kiwi Payments spend their day pulling data from product, payments, and marketing systems, writing SQL to answer business questions, building dashboards, and sharing insights that shape growth, retention, and fraud decisions. In fast-moving startups, you often work directly with product, growth, and risk teams, so you need to move quickly from a vague question to a clear, actionable answer. Expect a mix of deep-dive analysis, reporting, and cross-functional storytelling rather than just maintaining static reports.

Role and technical knowledge

1. 1

   How would you write a SQL query to find the daily active users for a UPI payments app over the last 30 days?

   A strong answer uses a date filter, counts distinct user IDs per day, and groups by date. Mention handling time zones and excluding test or internal accounts.

2. 2

   What is the difference between a left join and an inner join, and when would you use each in a payments dataset?

   Explain that left join keeps all rows from the left table, useful when you want all users even if they have no transactions. Inner join only keeps matches, useful when analysing only successful transactions.

3. 3

   Which metrics would you track to measure the health of a digital wallet or payments product?

   Cover activation rate, transaction success rate, average transaction value, repeat rate, and churn. Mention segmenting by user cohort or merchant type.

4. 4

   How would you explain a p-value or statistical significance to a non-technical product manager?

   Avoid jargon. Say it tells you how likely the result is due to chance, and connect it to whether an A/B test result is reliable enough to act on.

Practical scenarios

1. 5

   Transaction success rate dropped by 5 percent yesterday. How would you investigate?

   Break down by payment method, bank, device, app version, and geography. Check for recent releases, outages, or partner issues, and compare with the same day last week.

2. 6

   You have a dataset with missing values in the merchant category column. How would you handle it?

   First quantify how much is missing, check if it is random, and decide between dropping, imputing with a default like unknown, or using another column to infer. Explain the trade-offs.

3. 7

   The growth team wants to know if a new cashback offer increased repeat transactions. How would you measure it?

   Suggest an A/B test or a pre-post comparison with a control group. Define the metric clearly, check for seasonality, and mention sample size and duration.

4. 8

   A dashboard you built shows a sudden spike in new users, but the product team says sign-ups look normal. What could be wrong?

   Check for duplicate records, bot traffic, a tracking bug, or a change in how sign-ups are logged. Validate with a second source and communicate the uncertainty.

Business and product sense

1. 9

   How would you estimate the number of daily UPI transactions in a city like Pune?

   Break it down by population, smartphone penetration, UPI adoption, and average transactions per user per day. State assumptions clearly and sanity-check the final number.

2. 10

   A product manager asks you to find why users drop off during KYC. What data would you look at?

   Look at funnel steps, time spent per step, error messages, device type, and drop-off reasons. Suggest segmenting by acquisition channel and comparing with users who complete KYC.

3. 11

   How would you decide whether to show a new feature to all users or run an experiment first?

   Weigh the risk, cost, and reversibility. For a payments feature, mention compliance, fraud risk, and the need for a controlled test before a full rollout.

4. 12

   What is one metric you would not want to optimise blindly, and why?

   Pick something like average transaction value or number of transactions, and explain how it could hurt user trust, increase fraud, or reduce long-term retention.

Behavioural and HR

1. 13

   Tell me about a time you used data to solve a problem in a college project or internship.

   Use the STAR method. Focus on the question you asked, the data you used, the insight you found, and the decision or outcome it led to.

2. 14

   Why do you want to work at a fast-growing startup like Kiwi Payments instead of a large company?

   Show that you value ownership, speed, and learning across functions. Mention a specific interest in payments or fintech and how you handle ambiguity.

3. 15

   Describe a time you had to explain a technical finding to someone non-technical.

   Highlight how you simplified without losing accuracy, used visuals or analogies, and checked for understanding. Mention the outcome of the conversation.

4. 16

   Where do you see yourself in two years, and how does this role fit?

   Be honest but aligned. Show you want to deepen your analytics and business skills, take on more ownership, and grow with the company rather than jump ship quickly.