Remote data analyst interview questions and how to answer them
Remote data analyst interviews usually run in three or four stages: a recruiter call, a timed SQL test or take-home task, a technical interview where you talk through your work, and a final call about how you work with people. The questions below are the ones that come up again and again, with what the interviewer is checking and how to answer well.
Expect at least one live SQL question, one statistics question and a few about working without supervision across time zones.
Updated October 2026 by the WorkRemote careers team
On this page
About the role and your experience
Openers that check you understand the job and have done the work.
Walk me through one project from your portfolio.
Why they ask
It shows how you think from question to recommendation, and whether you actually did the work.
How to answer
Spend about two minutes. Give the business question, the data and its problems, what you did, the result in numbers, and what you'd do differently. Have the GitHub link or dashboard open in case they ask you to share your screen.
Show a sample answerHide the sample answer
I analysed two years of orders from a Brazilian e-commerce dataset to find why some customers never ordered again. Late deliveries were concentrated in a handful of states, and customers whose first order arrived late came back far less often than those whose order was on time. I cleaned duplicate customer IDs in SQL, built a cohort table and a Power BI dashboard, and recommended the company fix delivery estimates in three states first. If I did it again, I'd test whether price played a part too.
How would you define and measure churn for a subscription business?
Why they ask
Checks business sense. Churn sounds simple but has many definitions.
How to answer
Ask about the business first: monthly or annual plans, does a paused account count, customer churn or revenue churn. Then give a clear formula and say you'd agree it with finance.
Show a sample answerHide the sample answer
I'd first agree what counts as churned, say a customer whose subscription ended and wasn't renewed within 30 days. Monthly customer churn would be customers lost in the month divided by customers at the start of the month. I'd also track revenue churn, because losing one big client can matter more than losing ten small ones.
Why do you want this job, and why remote?
Why they ask
They want to know you understand the company and that remote work suits you, not just that you want dollars.
How to answer
Name something specific about their product or data, link it to your experience, and give a sensible reason remote suits you (focus, past remote or freelance work).
Show a sample answerHide the sample answer
Your product sells to small businesses across Africa and Europe, and I've spent the last year analysing SME payment data, so the questions you're solving are ones I already think about. I've worked remotely on Upwork contracts for two years and I'm good at managing my own time and writing clear updates.
Technical questions
What you know and how you'd do the work, often with a live task or test.
What is the difference between WHERE and HAVING?
Why they ask
A quick filter question. Getting it wrong suggests your SQL is shallow.
How to answer
WHERE filters rows before grouping; HAVING filters groups after aggregation. Give a one-line example.
Show a sample answerHide the sample answer
WHERE works on individual rows before GROUP BY, so I'd use it for
WHERE country = 'NG'. HAVING works on the grouped result, likeHAVING SUM(amount) > 10000to keep only customers who spent more than that.Write a query that returns the top three customers by revenue for each month.
Why they ask
Window functions separate people who have used SQL at work from people who finished one course.
How to answer
Aggregate first, then rank inside each month with a window function, then filter. Mention ties:
RANK()keeps tied customers,ROW_NUMBER()picks exactly three.Show a sample answerHide the sample answer
```sql
SELECT month, customer_id, revenue
FROM (
SELECT DATE_TRUNC('month', order_date) AS month,
customer_id,
SUM(amount) AS revenue,
RANK() OVER (PARTITION BY DATE_TRUNC('month', order_date)
ORDER BY SUM(amount) DESC) AS rnk
FROM orders
GROUP BY 1, 2
) t
WHERE rnk <= 3
ORDER BY month, rnk;
```After joining two tables, your row count went up. Why might that happen?
Why they ask
Join fan-out is one of the most common causes of wrong numbers on dashboards.
How to answer
Explain that the join key isn't unique in one table, so each row matches several rows. Say how you check (count distinct keys on each side before joining) and fix it (aggregate first, deduplicate, or join on a more specific key).
Show a sample answerHide the sample answer
Probably the key isn't unique on one side. If I join orders to payments and some orders have two payment attempts, each of those orders appears twice and revenue doubles. I'd compare
COUNT(*)withCOUNT(DISTINCT order_id)on the payments table, then either aggregate payments per order first or keep only the successful payment.What is the difference between an INNER JOIN and a LEFT JOIN, and when would a LEFT JOIN give you a misleading result?
Why they ask
Checks that you understand which rows survive a join, and the classic trap of filtering the right-hand table in WHERE.
How to answer
INNER keeps only matches; LEFT keeps every row from the left table and fills non-matches with NULL. Then mention that a condition on the right table in WHERE quietly turns a LEFT JOIN into an INNER JOIN; put it in the ON clause instead.
Show a sample answerHide the sample answer
If I want all customers, including those with no orders this month, I use a LEFT JOIN from customers to orders. But if I write
WHERE orders.order_date >= '2026-09-01', customers with no orders get dropped because their order_date is NULL. The date condition belongs in the ON clause.How do you handle missing or bad data?
Why they ask
Real company data is messy. They want to know you won't silently delete half of it.
How to answer
Say you first find out why it's missing, then decide case by case: fix it at source, fill it with a clear rule, exclude it and say so, or flag it. Always report how many rows were affected.
Show a sample answerHide the sample answer
First I check how much is missing and whether there's a pattern. On one project, 12% of orders had no state, and almost all came from one sales channel, so dropping them would have hidden that channel. I labelled them 'Unknown' and told the team so they could fix the form.
When would you report the median instead of the mean?
Why they ask
A basic statistics check that comes up constantly with revenue, salaries and delivery times.
How to answer
When the data is skewed or has outliers. Give a real example where the mean misleads.
Show a sample answerHide the sample answer
For order values. If most customers spend about ₦15,000 but a few corporate buyers spend ₦2 million, the mean jumps and describes nobody. The median shows the typical customer, and I'd show the big buyers separately.
An A/B test shows conversion of 4.1% for control and 4.4% for the new checkout, with 10,000 users in each group. Should we launch?
Why they ask
Tests whether you understand statistical significance and won't hand a manager a false win.
How to answer
Say you'd check significance before deciding, and mention sample size, test length and practical impact. Roughly: a two-proportion z-test here gives z of about 1.05 and a p-value near 0.29, so the difference could easily be noise.
Show a sample answerHide the sample answer
Not yet. With 10,000 users per group, a 0.3-point difference isn't statistically significant; I get a p-value around 0.29. I'd check the test ran for full weeks, look at whether the effect holds across devices, and work out how many users we'd need to detect a lift this small. If it's cheap to keep running, I'd keep running it.
In Power BI, what is the difference between a calculated column and a measure?
Why they ask
Common for Power BI roles; it shows whether you understand the data model or just drag fields around.
How to answer
A calculated column is computed row by row when data refreshes and stored in the model. A measure is computed at query time in the current filter context. Use measures for totals and ratios; use columns when you need to slice or filter by the value.
Show a sample answerHide the sample answer
A calculated column, like a customer's age band, is stored for every row and I can put it on an axis. A measure, like
Total Revenue = SUM(Sales[Amount]), is calculated for whatever filters the report has, so it gives the right number by month, region or product. I use CALCULATE when I need to change that filter context, for example revenue for last year.
Behavioural questions
Past situations. Answer with a real example: what happened, what you did, what changed.
Tell me about a time your analysis changed a decision.
Why they ask
They want proof your work has had impact, not just output.
How to answer
Use a short story: situation, what you found, what changed, the result. A project, a volunteer job or a previous non-data role is fine if you used data.
Show a sample answerHide the sample answer
At my last job in a Lagos logistics company, managers wanted to hire two more dispatch riders for Lekki. I pulled six months of delivery data and found the delays came from one pickup window, not rider numbers. They moved the pickup time instead, and late deliveries in that area fell over the next two months without new hires.
Tell me about a mistake you made in an analysis.
Why they ask
Everyone makes mistakes. They want to see honesty and the checks you built afterwards.
How to answer
Pick a real but recoverable mistake. Say how you found it, who you told, how you fixed it and what check you now run every time.
Show a sample answerHide the sample answer
I once reported a 20% rise in new users that was really a duplicate import. A colleague questioned it, I traced it within the hour and sent a correction to everyone who'd seen the report. Now I always compare row counts and distinct IDs against the previous load before I publish.
How would you explain a technical finding to someone who doesn't work with data?
Why they ask
Most stakeholders in remote teams read your summaries without you there to explain.
How to answer
Lead with the answer and what to do, use one clear chart, avoid jargon and give a sense of certainty in plain words.
Show a sample answerHide the sample answer
I'd say: "Customers who use the app in their first week are about twice as likely to still be with us after three months, so we should push first-week app use in onboarding." Then one chart, and a line on what we don't know yet.
Situational questions
Hypotheticals about problems you'd meet in the job.
Your dashboard shows revenue of $412,000 for September but finance says $398,000. What do you do?
Why they ask
Mismatched numbers happen weekly in real jobs. They want a calm, methodical person who protects trust in the data.
How to answer
Don't guess or defend your number. Explain how you'd reconcile: definitions (gross or net of refunds, booking date or payment date, currency conversion), filters, time zones on timestamps and duplicates. Tell the stakeholder early that you're checking.
Show a sample answerHide the sample answer
I'd tell finance and the dashboard users that I'm checking, so nobody quotes either number yet. Then I'd compare definitions: finance usually reports net of refunds and by payment date, while dashboards often use order date. I'd pull both at transaction level and find the rows that differ. Once I know the cause, I'd document the definition and fix whichever is wrong.
A manager asks you for 'a dashboard with everything on it'. How do you respond?
Why they ask
Shows whether you can push back politely and work out the real need.
How to answer
Say you'd ask what decisions the dashboard should support and who will use it, then propose a first version with a few key metrics and add more only when someone needs them.
Show a sample answerHide the sample answer
I'd ask what decisions they want to make with it and how often they'll look. Usually it comes down to four or five numbers. I'd build that first in a few days, then add views only when someone asks a question the dashboard can't answer.
Remote work questions
How you'll work from Nigeria with a team elsewhere: hours, power, internet, communication.
We're in London and some of the team is in New York. How would you manage your hours and communication?
Why they ask
They need to know you'll overlap with the team and won't go quiet.
How to answer
Be specific about hours, how you'll share progress and how fast you reply. Mention that Lagos and London share hours for most of the year.
Show a sample answerHide the sample answer
Lagos is one hour ahead of London in winter and on the same time in summer, so I'd work your core hours. For New York, I'd keep a couple of hours in my evening for calls a few days a week. I post a short update in Slack at the end of each day with what I finished, what's blocked and what's next, so nobody has to chase me.
What is your setup for power and internet?
Why they ask
Employers who've hired in Nigeria know power cuts are real. They want to hear you've planned for it.
How to answer
Be concrete: backup power, two internet sources from different providers, and what you'd do if both fail. Don't oversell.
Show a sample answerHide the sample answer
I have an inverter with batteries that runs my laptop and router for about eight hours, plus a small generator for longer outages. My main internet is a fibre connection and I keep a MiFi on a different network as backup. If both ever failed, I'd message the team immediately and work from a co-working space nearby.
How do you approach a take-home task with a deadline?
Why they ask
Take-homes are how many remote employers test analysts, and they want to see time management and judgement.
How to answer
Explain that you clarify the question, set a time limit, write down assumptions, focus on the two or three most useful findings and present them clearly rather than trying to analyse everything.
Show a sample answerHide the sample answer
I read the brief twice and email any questions in the first hour. I set myself the time they suggest, so if it's four hours I stop at four. I put my assumptions at the top, lead with the main finding and recommendation, and list what I'd look at with more time.
Remote interview tips
Test your setup the day before. Charge the inverter, check both internet connections and have your phone hotspot ready. Join video calls five minutes early.
Expect to share your screen. Have your portfolio, GitHub and a blank SQL editor (or a tool like DB Fiddle) open so you can write queries live without fumbling.
Talk while you code. In live SQL rounds, say what you're doing and why. Interviewers will forgive a syntax slip if your logic is clear, and they can't read your mind on a video call.
Confirm the time zone in writing. Write interview times in both zones ("2pm BST / 2pm WAT" or "10am EST / 4pm WAT") when you reply, especially around the March, October and November clock changes in the UK and US.
Follow up in writing. A short thank-you email that repeats one thing you discussed and answers anything you fumbled is normal in remote hiring and shows the written communication they're paying for.
Questions people ask
What SQL questions are asked in data analyst interviews?
Mostly joins, GROUP BY with HAVING, CASE WHEN, subqueries or CTEs, and window functions such as ROW_NUMBER, RANK and LAG. Expect a "top N per group" question and something with dates, like month-on-month growth. Many employers use timed tests on HackerRank, CodeSignal or TestGorilla.
How long is a data analyst take-home test?
Usually two to six hours of work, often with a few days to submit. If a task looks like more than a day's unpaid work, it's fair to ask whether they'd accept a shorter version or pay for it.
Do remote data analyst interviews include statistics?
Yes, usually at a basic level. Expect mean versus median, interpreting an A/B test, and correlation versus causation. Product analyst roles go deeper into experiment design.
What should I wear for a remote data analyst interview?
What you'd wear to an office interview from the waist up, in a quiet room with a plain background. Good light on your face and a working microphone matter more than the outfit.
Updated October 2026 by the WorkRemote careers team
Interview questions for roles close to data analyst
- Data Scientist
A data scientist uses statistics, machine learning and code (mostly Python and SQL) to predict outcomes and test what works, such as which loan applicants will default or whether a new feature lifts sales. You can do it remotely from Nigeria, but remote entry-level data science jobs are rare, so most people get in through a data analyst role, a strong portfolio or a master's degree.
- Data Engineer
A data engineer builds and runs the pipelines that move data from apps, payment systems and third-party tools into a warehouse where analysts and data scientists can use it. It is one of the stronger remote options for Nigerian engineers, because the work is cloud-based, skills are in short supply and most people move into it from backend development or analytics.
- Business Analyst
A business analyst works out what a business actually needs from a system or process change and writes it up as clear requirements that developers can build and test. It's a common remote role for Nigerians, especially those with banking, fintech or telecoms experience, though fully remote junior roles are scarce.
- Financial Analyst
A financial analyst builds budgets, forecasts and models that help a company decide where to spend money, and explains why results differ from the plan. Remote financial analyst work from Nigeria is realistic in FP&A (financial planning and analysis) roles at tech companies and startups; investment and equity research roles are far harder to get remotely.