Session 34 · Phase 5: Business & Mock Rounds

Mock Technical Interview

The one-on-one round: explaining concepts out loud, walking through your approach, and showing how you think — not just what you memorized.

⏱ ~1.5 hrs 🧠 Mock 🎯 Highest priority

1. What the Verbal Technical Interview Really Tests

The written assessment checks what you know; the verbal interview checks how you think. Interviewers are looking for:

💡
The #1 rule: never answer with one word. Explain what, why, and a quick example. "It's a join" is weak; "a LEFT JOIN keeps all left rows, so I'd use it to find customers with no orders" is strong.

2. "What Is the Data Analysis Process?"

This is asked constantly. Have the five steps ready:

  1. Collect — pull data from databases/files.
  2. Clean — handle missing values, duplicates, inconsistencies.
  3. Analyze — find trends, patterns, anomalies.
  4. Visualize — build charts and dashboards.
  5. Recommend — turn findings into an action.

3. Mock Interview Q&A — Practice Out Loud

Answer each aloud before revealing. Record yourself if you can.

1. Tell me about yourself.

Model (1 min): "I'm a data analyst with strong SQL, Excel, and Python skills. I've built a churn analysis in Python and a sales dashboard in Power BI. I enjoy turning messy data into clear insights, and I'm looking to apply that in a banking analytics role."

2. What is a VLOOKUP, and what are its limitations?

Model answer: "VLOOKUP searches down the first column and returns a value to its right. Its limits: it can only look right, the key must be the first column, and it defaults to approximate match. XLOOKUP or INDEX-MATCH fixes these."

3. What is a pivot table, and how do you make one?

Model answer: "A pivot table summarizes data by category without formulas. I select the data, go to Insert → PivotTable, and drag fields into Rows, Columns, and Values."

4. How do you handle missing or duplicate data?

Model answer: "For duplicates, I remove them. For missing data, I first understand why it's missing, then drop or fill it. In Pandas that's drop_duplicates(), dropna(), and fillna()."

5. What's the difference between WHERE and HAVING?

Model answer: "WHERE filters rows before grouping; HAVING filters groups after aggregation. So HAVING works on things like SUM or COUNT, WHERE doesn't."

6. Tell me about a challenging data problem you solved.

Model (STAR): "Situation: churn was rising. Task: find drivers. Action: I cleaned and segmented the data. Result: I found single-product inactive customers churn most, and recommended a retention campaign."

7. What are your strengths as a data analyst?

Model answer: "I'm strong at SQL and cleaning messy data, and I communicate findings in plain English — I've learned an insight only matters if a non-technical stakeholder can act on it."

8. Do you have any questions for me?

Model answer: "Yes — what does a typical day look like for the team, and what does success look like in this role in the first six months? I'd also love to know what tools the team uses day to day."

📋 Stable content — Reviewed: August 2026

4. Interview Questions (with Model Answers)

The verbal technical questions interviewers ask most. Self-test before revealing.

IQ1. What is the data analysis process?

Model answer: "Collect, clean, analyze, visualize, and recommend. I always start by clarifying the business question, then clean the data, find patterns, and end with a recommendation."

IQ2. How do data analysts differ from data scientists?

Model answer: "Analysts focus on describing the past and present — reports, dashboards, and insights. Data scientists focus more on predictive modeling and machine learning. There's overlap, but analysts lean toward business reporting."

IQ3. What software/tools have you used for analysis?

Model answer: "SQL for querying, Excel for quick analysis and pivot tables, Python (Pandas/NumPy) for cleaning and EDA, and Power BI/Tableau for dashboards."

IQ4. How do you explain a technical result to a non-technical audience?

Model answer: "I lead with the takeaway in plain language, use a simple chart, avoid jargon, and tie it to a decision they care about — 'this segment is churning, so we should target them with this offer.'"

IQ5. What do you do when you get unexpected results?

Model answer: "First I check for a data or logic error. If the result is real, I dig in — segment the data to understand what's driving it. Unexpected results are often where the real insight is."

IQ6. What's the difference between a measure and a calculated column in Power BI?

Model answer: "A measure is computed dynamically and not stored; a calculated column is stored per row. I use measures for aggregations like totals and ratios."

IQ7. How do you validate that your analysis is correct?

Model answer: "I cross-check totals against the source data, sanity-check the magnitude (does this number make sense?), and have someone else review it. I also re-run the logic on a subset I can verify by hand."

IQ8. How do you handle a question you don't know the answer to?

Model answer: "I'd say honestly that I'm not certain, then talk through how I'd figure it out — what I'd look up, or how I'd break the problem down. I'd rather show my reasoning than bluff."

Key Takeaways

1

The verbal round tests how you think, not just what you know.

2

Never answer with one word — give what, why, and an example.

3

Have the five-step data analysis process memorized.

4

Use STAR for project stories; lead with the business outcome.

5

Be honest about what you don't know — show reasoning instead of bluffing.

Objective Questions — Test Your Understanding

Q1. "Tell me about yourself" should focus on…

Q2. What is the correct order of the data analysis process?

Q3. What is a key limitation of VLOOKUP?

Q4. How should you answer a question you don't know?

Q5. What is a pivot table primarily used for?