Free Starter kit · Indian School of Skills

Data Analyst Starter Kit: Projects & Interview Questions

Last reviewed: September 2026

Entry-level data analyst interviews test three things: can you get and clean data, can you answer a business question with it, and can you explain the answer to someone who isn't technical. This kit gives you projects that prove all three, and the questions you should be ready for.

1. Skills checklist

Excel / Google Sheets

  • Cleaning: TRIM, text-to-columns, removing duplicates, fixing data types
  • Lookups: XLOOKUP (or INDEX/MATCH), and when VLOOKUP breaks
  • Logic and aggregation: IF/IFS, SUMIFS, COUNTIFS, AVERAGEIFS
  • Pivot tables and pivot charts
  • Basic charts that follow good practice (no 3-D pies)

SQL

  • SELECT, WHERE, ORDER BY, LIMIT
  • GROUP BY with HAVING; COUNT(DISTINCT)
  • INNER / LEFT joins, and spotting duplicate rows caused by joins
  • CASE WHEN for bucketing
  • Subqueries and CTEs (WITH)
  • Window functions: ROW_NUMBER, RANK, LAG/LEAD, running totals
  • Date handling: truncating to month, date differences

Power BI (or Tableau)

  • Power Query: import, clean and transform steps
  • Data modelling: star schema, fact vs dimension tables, relationships
  • DAX basics: measures vs calculated columns, CALCULATE, SUM/AVERAGE, time intelligence (e.g. TOTALYTD)
  • Slicers, drill-through and tooltips
  • Dashboard layout: KPI cards on top, trend, breakdown, detail

Statistics and thinking

  • Mean vs median, spread, outliers
  • Correlation vs causation
  • Sampling, and why small samples mislead
  • A/B test basics: control, variant, significance in plain words
  • Framing: turning a vague request into a specific question and metric

2. Five portfolio project briefs

For each project: write the business question first, keep a short log of cleaning steps, and finish with three recommendations a manager could act on. Always check a dataset's licence and description on its source page before using it.

Project 1 — Retail sales performance dashboard (Excel → Power BI)

Data: the "Sample - Superstore" dataset (widely available with Tableau and on Kaggle) or the UCI Online Retail dataset.

Business question: which regions, categories and customers drive profit, and where are discounts destroying margin?

Deliverables: cleaned data, a star-schema model, a one-page dashboard with KPIs (sales, profit, margin, orders), and a 5-bullet summary.

Project 2 — E-commerce funnel and delivery analysis (SQL)

Data: the Brazilian E-Commerce Public Dataset by Olist (Kaggle): orders, items, payments, reviews and customers in separate tables.

Business question: how do late deliveries affect review scores, and which states or sellers are most affected?

Deliverables: 8–10 SQL queries (with joins, CTEs and a window function), a short write-up and one chart per finding.

Project 3 — Customer cohort and retention analysis

Data: UCI Online Retail II (transaction-level data with invoice dates and customer IDs).

Business question: what share of customers come back in months 1, 2 and 3 after their first purchase, and did any cohort behave differently?

Deliverables: a cohort retention table/heatmap, an RFM segmentation (recency, frequency, monetary) and suggested actions for each segment.

Project 4 — Public data story for India

Data: a dataset from data.gov.in (Open Government Data Platform India), e.g. on agriculture, health, education or transport.

Business question: choose one clear question (e.g. "how has X changed across states over the last decade?").

Deliverables: a cleaned dataset, a dashboard with a map or ranked bar chart, and a 300-word narrative with caveats about data quality.

Project 5 — Marketing campaign / A/B test analysis

Data: a public marketing A/B test or campaign dataset from Kaggle (search "A/B testing" or "marketing campaign").

Business question: did the variant perform better, by how much, and is the difference likely to be real?

Deliverables: conversion rates per group, a significance test explained in plain language, and a recommendation with its risks.

3. Interview questions to prepare

SQL

  1. What is the difference between WHERE and HAVING?
  2. Explain INNER, LEFT, RIGHT and FULL joins. When would a LEFT join return more rows than the left table?
  3. Find the second-highest salary in a table (show two methods).
  4. Find customers who ordered in January but not in February.
  5. What is the difference between RANK, DENSE_RANK and ROW_NUMBER?
  6. Calculate a 7-day rolling average of daily sales.
  7. How would you find and remove duplicate rows?
  8. What is a CTE and why would you use one instead of a subquery?

Excel and Power BI

  1. XLOOKUP vs VLOOKUP: what are VLOOKUP's limitations?
  2. What is the difference between a measure and a calculated column in Power BI?
  3. What does CALCULATE do? Give an example.
  4. What is a star schema and why does it matter for performance and correct totals?
  5. How would you handle a many-to-many relationship?
  6. How do you decide which chart to use?

Statistics and business sense

  1. When would you report the median instead of the mean?
  2. Sales dropped 15% last week. How would you investigate?
  3. How would you define and measure "active user" for an app?
  4. Explain a p-value to a non-technical manager.
  5. What is the difference between correlation and causation? Give an example.
  6. How do you handle missing values and outliers?

Behavioural

  1. Tell me about a time your analysis changed a decision.
  2. How do you explain a technical finding to a non-technical audience?
  3. Tell me about a time the data was messy or wrong. What did you do?
  4. How do you prioritise several urgent data requests?

4. Present a project in 3 minutes

  1. Question (20 sec): the business question and why it matters.
  2. Data (30 sec): source, size, and the one cleaning issue that mattered most.
  3. Approach (40 sec): the analysis and tools, briefly.
  4. Findings (60 sec): the top two or three insights, each with one chart.
  5. Recommendation (30 sec): what you would do, and what you'd check next.

5. Résumé bullet formula

Action verb + what you analysed + tool + result or insight. For example: "Analysed 100k+ e-commerce orders in SQL to show that late deliveries were linked to lower review scores; recommended seller-level SLA alerts." Only claim numbers that are true for your project.

Want structured feedback on this?

The ISS Data & Business Intelligence program is a live online cohort where mentors critique the projects you build. Talk to an advisor to see whether it fits your goals, or explore the curriculum first.

View the Data & Business Intelligence curriculum → Ask an advisor on WhatsApp

© Indian School of Skills. You're welcome to print and share this resource for personal learning. Tools, fees and third-party programme details change; verify current information with each provider.