Weak Answer (Most Freshers):
Q: Tell me about your data analysis work.
A: I know Excel, SQL and Power BI. I made a sales dashboard.
Problem: No business context, no logic, no outcome
Result: Interviewer moves on → rejection

Selection-Ready Answer:
Q: Tell me about your data analysis work.
A: One region's revenue dropped 12%. I pulled 80K sales rows
   from 3 tables using SQL JOINs and window functions,
   cleaned duplicates in Excel, and built a Power BI KPI view.
   Finding: stock-outs in the top 3 SKUs caused the drop.
   Recommendation: inventory reallocation → stock-outs fell 18%.
Result: Interviewer asks deeper questions → next round

What the Interviewer Scores:
1. Technical accuracy  - is the query/logic correct?
2. Approach            - did you structure the problem first?
3. Data handling       - cleaning, assumptions, edge cases
4. Business sense      - does the answer help a decision?
5. Communication       - simple, structured, confident
6. Depth of projects   - can you defend what you built?
Tool names alone score almost nothing.

Most Asked Practical Questions:
- Difference between INNER JOIN and LEFT JOIN (with example)
- Find the second highest salary / top 3 products per region
- Remove duplicate rows and explain why they occurred
- GROUP BY vs window function - when to use which
- Handle NULL values in aggregation
- Excel: VLOOKUP vs INDEX-MATCH vs XLOOKUP
- Power BI: measure vs calculated column, basic DAX
- Which chart for trend, comparison, and distribution?

Round-wise Preparation Focus:
Round 1 - Resume    : targeted skills + measurable projects
Round 2 - Screening : clear intro, why data analytics
Round 3 - Technical : SQL logic, Excel cleaning, BI basics
Round 4 - Case      : structure the problem before solving
Round 5 - Final     : ownership, communication, culture fit

Rejection reason #1 for freshers:
Correct query, but no explanation of the business meaning.

Case Study Framework (Use Out Loud):
1. Clarify  - What exactly are we trying to measure?
2. Metrics  - Which KPIs answer this question?
3. Data     - Which tables, what time period, any gaps?
4. Analysis - Segment by region/product/customer/time
5. Insight  - What changed, where, and by how much?
6. Action   - What should the business do next?
7. Risk     - What assumptions could break this?

Speaking your steps aloud scores more than a silent answer.

Weak Answer:
- I cleaned the data and made a dashboard in Power BI.

Strong Answer:
- I worked on 80K sales rows across 3 tables for 18 months.
  Removed 4% duplicate order IDs, treated NULL discounts as zero,
  then built a Power BI view by region and SKU.
  Insight: 12% revenue drop came from stock-outs in 3 SKUs.
  Action: reallocation plan that reduced stock-outs by 18%.
  Assumption: returns data was not available for 2 months.

Communication Checklist (Scored in Every Round):
- Answer the business question first, detail second
- Use numbers, not vague words like "improved a lot"
- Explain the insight as if the listener is non-technical
- State assumptions before conclusions
- Say "I don't know, but here is how I would find out"
- Keep each answer under 90 seconds
- Ask one clarifying question before solving a case

Data Analyst Interview - 7 Day Prep Plan:

Day 1-2: SQL
- JOINs, GROUP BY, HAVING, subqueries, window functions
- Write queries by hand, then verify

Day 3: Excel + data cleaning
- Duplicates, NULLs, pivot tables, lookup functions

Day 4: Power BI / visualization
- KPI selection, chart choice, basic DAX measures

Day 5: Projects
- Rehearse each project: problem, data, insight, impact

Day 6: Case studies
- Solve 2 business cases out loud using a framework

Day 7: Mock interview
- Full round + feedback + fix weak answers

Helpful Resources:

Free:
- Public datasets for SQL, Excel and Power BI practice
- SQL practice platforms with real interview problems
- Company-wise interview experience blogs

Paid / Structured:
- Uncodemy - Data Analytics Training Course
- 1:1 mock interviews with trainer feedback
- Portfolio and project review sessions

Certifications (Supporting only):
- Useful only when paired with real projects
- A certificate alone never clears a technical round