Data Analyst Interview Practice: Take-Home Tasks, SQL-Style Questions, Metric Definitions, Quick A/B Reads & Sanity Checks

What you'll build

What you will be able to do

Each lab ends with a concrete result that you build yourself and the checker confirms, cell by cell.

• Answer the classic warm-up questions precisely: a change in percentage points and in percent, a channel's share of revenue and its share of the growth.
• Answer three SQL questions without SQL: a LEFT JOIN of orders to customers with a label for unmatched rows, a count of orders per customer including customers with none, and revenue per region.
• With one small user table you calculate weekly active users under two definitions and conversion with two denominators – and see how far the answers move.
• Describe 20 basket values with the mean, the median and the 90th percentile, then show how one large business order moves the mean but hardly the median.
• Read a finished checkout test like an interviewer expects: conversion per variant, the absolute lift in points, the smallest change the test can detect from a fictional lookup table, and a verdict.
• Turn an activity log into a cohort table: months since signup per row, active users per cohort and month, retention percentages and the pooled month-1 retention.
• Audit a daily export before using it: flag duplicated dates, count impossible days, find rows whose reported conversion is wrong, and explain the gap to the dashboard total.
• Against the clock you answer a mini take-home: net revenue, the refund rate, net revenue per channel and a chart that a hiring manager can read in five seconds.