Data analyst interviews test whether you can turn messy data into a clear answer and explain it to people who must act on it. Expect questions about SQL, cleaning data, choosing measures and presenting findings. To prepare, pick two or three projects from your work, and practise explaining the question, the method and the decision that followed.
Ten questions, with sample answers
The sample answers are original examples. Change the details to match your own work, and never claim tools or results that are not yours.
1. Walk me through an analysis from start to finish.
What they look for: that you start with the question, not the data.
Sample answer: “A manager asked why orders dropped in one region. I clarified what counted as a drop, pulled the data, cleaned it, compared periods and checked a few explanations. I shared a short summary with one chart and a recommended next step.”
2. How do you clean a messy dataset?
What they look for: a careful, documented routine.
Sample answer: “I check for duplicates, missing values and impossible entries, and decide case by case whether to fix, remove or flag them. I keep a log of every change so someone else can repeat the work.”
3. Explain the difference between an inner join and a left join.
What they look for: a precise explanation with a quick example.
Sample answer: “An inner join keeps only rows with a match in both tables. A left join keeps every row from the left table and fills the right side with empty values where there is no match. I use a left join when I must not lose customers who have no orders.”
4. How do you choose which metric to report?
What they look for: linking measures to decisions.
Sample answer: “I ask what decision the number will support. Then I pick a measure that moves when that thing changes, define it clearly and agree it with the people who will use it.”
5. Tell me about a time your analysis changed a decision.
What they look for: impact, and a clear role.
Sample answer: “My comparison showed that a popular report was counting some customers twice. I shared the evidence, we fixed the definition, and the team changed its plan based on the corrected figures.”
6. How do you explain results to a non-technical audience?
What they look for: clarity and focus.
Sample answer: “I lead with the answer in one sentence, show one simple chart and keep the method for those who ask. I check that they know what to do next.”
7. What would you do if the numbers look too good to be true?
What they look for: healthy scepticism.
Sample answer: “I would check the source, the filters and the date range, look for duplicates and compare with an independent figure. I would only share the result once I could explain it.”
8. How do you deal with a stakeholder who keeps changing the request?
What they look for: patience and structure.
Sample answer: “I write down the question and the definition of success, and agree it before I start. If it changes, I explain what that does to the timing and confirm the new version in writing.”
9. Which tools do you use, and how do you choose?
What they look for: honest skills and good reasons.
Sample answer: “I use SQL to get and shape data, a spreadsheet for quick checks and [your chart tool] for sharing. I pick the simplest tool that is reliable and that my team can maintain.”
10. Why are you interested in this role?
What they look for: curiosity about their data and problems.
Sample answer: “The post describes working closely with product and operations teams, which is where I do my best work. I enjoy turning an open question into something a team can act on.”
How to prepare
- Revise your core skills. Practise joins, grouping and filtering in SQL, and be ready to say what each query returns.
- Prepare two project stories. For each, state the question, the data, the method, the finding and the decision it supported.
- Show your checking. Interviewers like to hear how you tested for duplicates, gaps and odd values before trusting a result.
- Practise aloud. Explaining a method clearly is harder than doing it, so say it out loud before the day.
How to practise this with PrepHint AI
Every account starts with 15 free minutes of live help and 1 mock interview of up to 10 minutes. No card needed.
- Add your resume. Sign up at PrepHint AI, then add your resume, or build one first in Resume Maker, which is free forever with an account.
- Create a mock interview for this job. Add the job post, choose your resume and a length from 5 to 30 minutes. The questions are built from the job post and your resume.
- Practise out loud by voice. The AI interviewer asks its questions aloud and listens to your spoken answers. Practise explaining a method in plain words.
- Read the transcript. The transcript of every mock interview is saved, so you can read your own answers afterwards and cut the waffle.
- For live answers in a call, use live answers in the app.
Use it where AI help is allowed. PrepHint AI is made for practice and for conversations where help from AI is allowed. Check the rules of the employer and the interviewer before you use it in a real interview. If the answer is no, use it to practise.
Read more about the AI mock interview or see the plans on the pricing page.