Skip to content

Interview Questions — Introduction to Data Analytics

Conceptual interview questions on analytics fundamentals. These test whether you understand analytics as a business discipline, not just a set of tools.

Fundamentals

[Beginner] What is data analytics, and how is it different from data science?

Show answer

Data analytics is the practice of examining data to answer business questions and support decisions. It focuses heavily on descriptive and diagnostic work — what happened and why — using SQL, Excel, and BI tools.

Data science extends into predictive and prescriptive territory — building machine learning models, forecasting, and optimisation — using more advanced statistics and programming.

The honest distinction: analysts answer "what happened and why, and what should we do about it" using established methods; data scientists build the predictive systems and custom models. In practice the roles overlap, and many analysts do light predictive work.

[Beginner] What are the four types of analytics?

Show answer
  1. Descriptive — "What happened?" (reports, dashboards, summaries)
  2. Diagnostic — "Why did it happen?" (root cause, drill-down, attribution)
  3. Predictive — "What will happen?" (forecasting, churn models)
  4. Prescriptive — "What should we do?" (optimisation, A/B testing, decision support)

Each increases in complexity and value. Most analyst work is descriptive and diagnostic, with selective predictive work. Prescriptive typically involves data scientists.

[Beginner] What does a data analyst actually do day to day?

Show answer

A realistic answer (not just "analyse data"): - Talk to stakeholders to understand what decision they're trying to make - Pull and clean data (often the biggest time sink — 60-80%) - Run analysis in SQL, Python, or Excel - Build dashboards and reports - Present findings and recommendations - Answer follow-up questions and iterate

The non-obvious truth: the technical work is maybe half the job. The other half is communication — understanding the question and presenting the answer so it drives a decision.

Applied Thinking

[Mid-level] A stakeholder asks for "a dashboard with all our data on it." How do you respond?

Show answer

I wouldn't just build it — a dashboard with everything communicates nothing. I'd dig into the real need:

  • "What decisions are you trying to make with this dashboard?"
  • "What would you check first thing in the morning?"
  • "What would make you take action?"

Then I'd propose a focused dashboard of 5-7 KPIs tied to their actual decisions, with the option to drill into detail. The goal is to replace "all our data" with "the data you act on."

This shows I understand that analytics serves decisions, not data completeness.

[Mid-level] What's the difference between an interesting insight and an actionable one?

Show answer

An interesting insight is true but doesn't change what anyone does. "Sales are higher on Tuesdays" is interesting — but if nobody can act on it, it's trivia.

An actionable insight points to a specific decision or change. "Sales are higher on Tuesdays because that's when our email campaign sends — and Thursday sends underperform by 40%, so we should move all sends to Tuesday" is actionable.

The test: can someone do something differently because of this? If not, it's documentation, not analysis. Good analysts ruthlessly filter for actionability.

[Mid-level] Walk me through how you'd approach a request to "find out why sales dropped last month."

Show answer

This is diagnostic analytics. My approach:

  1. Clarify: Dropped compared to what — last month, last year, forecast? How big is the drop? Is it one product/region or across the board?
  2. Segment: Break the drop down by dimension — product, region, channel, customer segment, time. The drop is usually concentrated somewhere.
  3. Isolate: Find where the decline is concentrated (e.g., "it's entirely in the West region's Electronics category").
  4. Investigate the cause: Did pricing change? A competitor launch? A stockout? A data issue?
  5. Recommend: Based on the cause, what should we do?

I'd avoid jumping to conclusions — segment first, then form a hypothesis, then verify.

Senior / Strategic

[Senior] How do you measure the impact or value of an analytics team?

Show answer

Not by output volume (number of dashboards or reports) — that's a vanity metric. Analytics value is measured by decisions influenced and outcomes improved:

  • Decisions made differently because of analysis (and their business impact)
  • Time saved through automation (e.g., a dashboard replacing manual reporting)
  • Revenue gained or cost avoided from insights acted upon
  • Speed-to-insight — how fast can the business get answers?

The hardest and most honest measure: would the business make worse decisions without the team? I'd also track adoption (are dashboards actually used?) as a leading indicator of value.

[Senior] What organisational factors cause analytics to fail to create impact?

Show answer

Common anti-patterns:

  1. No decision owner — analysis produced with no one accountable for acting on it
  2. Order-taking culture — analysts build what's requested without questioning whether it answers the real need
  3. Data trust issues — if stakeholders don't trust the numbers, they ignore them; reconciliation and governance matter
  4. Insight-action gap — findings presented but no mechanism to turn them into decisions
  5. Vanity metrics — measuring activity (reports produced) instead of outcomes (decisions improved)
  6. Poor communication — brilliant analysis lost in jargon or data dumps

The fix is cultural as much as technical: embed analysts in decision-making, frame work around decisions, and measure impact not output.

Previous: 06-practice-exercises | Next: Excel for Analytics