- Karina Datascientist's Newsletter
- Posts
- Your First Data Interview: What To Expect
Your First Data Interview: What To Expect
Karina Samsonova
November 14, 2025
Interviews are stressful.
You're sitting there wondering: Will they ask me to write a SQL query while I am sharing my screen? Do I need to know every pandas function? What if they ask about statistical concepts and I blank?
Today, let's talk about what actually happens in data analyst interviews (from my personal experience as both interviewer and interviewee) — and how to prepare.
After conducting dozens of interviews at my consultancy and going through plenty myself, here's the truth:
Most interviewers aren't trying to trick you. Although, I do think that some are indeed trying to trick you. They're trying to see if you can actually do the job.
The questions fall into a few predictable categories. Let's break them down.
Category 1: The "Walk Me Through" Questions
These test how you think through problems.
"Walk me through how you would analyse [business problem]."
Example: "Our email open rates dropped 15% last month. How would you investigate?"
What they're really testing: Your analytical thinking process, not whether you know the "right" answer.
How to answer:
Clarify the problem ("What changed last month? New email provider? Different audience?")
Identify what data you'd need ("Email metrics, subscriber data, timing...")
Outline your approach ("First I'd check... then I'd compare... then I'd look for...")
Mention limitations ("I'd also want to know...")
Don't: Jump straight to solutions. Show your thinking process.
"How would you approach cleaning a messy dataset?"
What they're testing: Do you understand that data quality matters?
Good answer framework:
Check for obvious issues (nulls, duplicates, data types)
Understand WHY the data is messy (bad source? Integration issues?)
Validate against business logic (does this number make sense?)
Document what you changed and why
Category 2: The Technical Questions
These test if you actually know the tools. When I interview people I allow them to use any extra tools - google search, ChatGPT, call a friend. Because when I’m stuck on some task - this is exactly what I do. For me, it is important to see that you are comfortable with the tool, and if you are stuck - you will find a solution. However, I personally have been in several interviews where they were recording the screen and you were not allowed to even use google to double check the syntax of ROW_NUMBER.
"Write a SQL query to [specific problem]."
Common ones:
Find top 10 customers by revenue
Calculate month-over-month growth
Identify duplicate records
JOIN multiple tables
How to prepare: Practice common patterns on LeetCode (Easy/Medium problems). You don't need to memorise syntax—you need to understand JOINs, GROUP BY, and basic aggregations.
"What's the difference between [technical concept A] and [technical concept B]?"
Common ones:
LEFT JOIN vs INNER JOIN
Mean vs Median (and when to use which)
COUNT vs COUNT DISTINCT
WHERE vs HAVING
What they're testing: Do you understand the fundamentals?
How to answer: Simple explanation + when you'd use each one. Real example is bonus points.
Category 3: The Behavioural Questions
These test if you're actually pleasant to work with and if you can fit in with the team. This is more important than you might think. A team consists of various personalities. Nobody wants conflicts at the workplace, so managers are trying to find people who will work well together.
I also always looking for people who can fill in weaknesses in the team. For example, I am strong at A, B, C, but week in D. So if the person I interview is strong at D - that’s perfect, because together we will become unstoppable. I am looking for this balance.
"Tell me about a time you had to explain technical findings to a non-technical audience."
What they're really asking: Can you communicate? Will our stakeholders understand you?
Framework (STAR method):
Situation: Brief context
Task: What you needed to do
Action: What you actually did (be specific)
Result: What happened (quantify if possible)
Example: "At my consulting role, I had to present churn analysis to executives who didn't know SQL. I created a simple dashboard showing just three key metrics and used the analogy of a leaky bucket to explain retention. They approved the $50K retention campaign based on that presentation."
Other common behavioural questions (all of these I’ve been asked many times):
"Tell me about a time your analysis was wrong." (They want to see you can admit mistakes and learn).
"How do you prioritise when you have multiple urgent requests?" (They want to see you can manage expectations)
"Describe a time you disagreed with a stakeholder." (They want to see you can push back diplomatically)
Pro tip: Have 3-4 stories ready that you can adapt to different questions.
Category 4: The Case Study/Take-Home
Some companies give you data and ask you to analyse it.
What they're testing: Can you actually do the work?
What they care about:
Did you answer the business question?
Is your analysis logical?
Can you explain your findings clearly?
Did you document your work?
What they DON'T care about:
Complex techniques
Perfect code
My advice: Treat it like a real work project. Include:
Executive summary (your answer in 2-3 sentences)
Your approach (what you did and why)
Key findings (with visuals)
Limitations (what you couldn't answer)
Recommendations (what should they do with this info)
The Questions YOU Should Ask
The interview goes both ways. Ask about:
About the role:
"What does a typical project look like?"
"Who would I work with most closely?"
"What tools and data sources would I use?"
About success:
"What would success look like in the first 90 days?"
"What are the biggest data challenges the team faces?"
About the team:
"How is the data team structured?"
"What's the process for requests and prioritisation?"
Red flags to watch for:
They can't explain what you'd actually be doing
No clear stakeholders or users for your work
Unrealistic expectations ("We need someone to build everything from scratch in 3 months")
The Mistakes That Kill Your Chances
❌ Saying "I don't know" and stopping. Say "I don't know, but here's how I'd figure it out". It is ok not to know something, but you can always say ‘I haven’t worked with this specific tool, but I worked with…’
❌ Trash-talking previous employers
❌ Pretending to know something you don't
The Things That Impress
✅ Admitting what you don't know, then explaining how you'd learn
✅ Asking clarifying questions before answering
✅ Connecting your answers to business impact
✅ Having opinions but staying flexible
The Bottom Line
You don't need to be perfect. You need to be competent, curious, and pleasant.
Prepare the common questions. Practice your stories. Get comfortable saying "I don't know, but here's how I'd approach it."
And remember: they're interviewing multiple people. Your job is to be memorable for the right reasons—clear thinking, good communication, genuine interest, enthusiasm.
You've got this.
Keep pushing 💪
Karina
Python tip
Use ? and ?? to see documentation and source code in Jupyter.
? = documentation
?? = source code
Grab your freebies if you haven’t done already:
Data Playbook (CV template, Books on Data Analytics and Data Science, Examples of portfolio projects)
Need more help?
Just starting with Python? Wondering if programming is for you?
Master key data analysis tasks like cleaning, filtering, pivot and grouping data using Pandas, and learn how to present your insights visually with Matplotlib with ‘Data Analysis with Python’ masterclass.
Building your portfolio?
Grab the Complete EDA Portfolio Project — a full e-commerce analysis (ShopTrend 2024) with Python notebook, realistic dataset, portfolio template, and step-by-step workflow. See exactly how to structure professional portfolio projects.
Grab your Pandas CheatSheet here. Everything you need to know about Pandas - from file operations to visualisations in one place.
 | Data Analyst & Data Scientist |