Data Analyst Interview Prep: SQL, Excel, Power BI, Statistics, Case Studies & Salary Negotiation

I still remember how different a data interview feels compared with a normal job interview.

You can spend weeks preparing definitions, memorize SQL syntax, and watch hours of Power BI tutorials. Then the interviewer gives you a messy dataset and asks something simple like:

“What would you do with this data?”

That is usually where the real test begins.

Data-related interviews aren’t only about whether you know SQL commands or can build a dashboard. Interviewers want to see how you think, whether you notice problems in the data, how clearly you explain your findings, and whether you can turn numbers into something a business can actually use.

I’ve found that the best preparation isn’t memorizing 100 interview questions. It’s practicing the kind of thinking you’ll need when the question isn’t exactly what you expected.

Here’s how I’d prepare.

Top SQL & Excel Questions Asked

SQL and Excel remain two of the most common practical skills tested in data roles.

The exact questions vary by company, but the underlying concepts tend to repeat.

SQL questions you should be ready for

Expect questions around:

• SELECT, WHERE, and ORDER BY
• GROUP BY and aggregate functions
• JOIN operations
• Subqueries
• Common table expressions (CTEs)
• Window functions
• CASE WHEN
• Date calculations
• Handling NULL values
• Duplicate records
• Finding the highest or lowest values
• Ranking customers or products
• Calculating percentages and averages

A common interview exercise might look something like this:

“Find the top three customers by total revenue.”

The interviewer isn’t necessarily interested in whether you can remember one particular query. They may want to see whether you understand the table relationships, aggregation, sorting, and limitations of the data.

Another common type of question is:

“What’s the difference between an INNER JOIN and a LEFT JOIN?”

Don’t just memorize the textbook definition.

Explain it practically.

An INNER JOIN returns records that match in both tables. A LEFT JOIN keeps all records from the left table even if there’s no matching record on the right.

Then explain when you’d use one.

That second part often makes your answer stronger.

Excel questions

For Excel, prepare for:

• XLOOKUP
• INDEX and MATCH
• SUMIFS
• COUNTIFS
• IF
• PivotTables
• Conditional formatting
• Removing duplicates
• Data cleaning
• Text functions
• Date calculations
• Basic charts
• Filtering and sorting
• Error handling

You may also receive an actual spreadsheet and be asked to investigate it.

This is where practical experience matters.

Don’t spend all your preparation time memorizing formulas. Practice taking a messy spreadsheet and answering a business question with it.

For example:

“Which products generated the most revenue last quarter, and which ones are declining?”

That requires more than a formula.

You need to understand the question, clean the data, calculate the right metrics, and explain what the numbers mean.

Real Dataset Analysis Problems

This is one of the most valuable areas to practice.

Instead of solving isolated questions, download a public dataset and pretend you’re already working for the company.

Kaggle is useful for this because you can find datasets covering sales, customers, marketing, finance, healthcare, e-commerce, and many other areas.

Don’t immediately start making charts.

Start with questions.

Imagine you’re analyzing an e-commerce dataset.

Ask:

• Which products generate the most revenue?
• Which customers purchase most frequently?
• Are sales growing or declining?
• Which regions perform best?
• Are there seasonal patterns?
• Which products have high sales but low margins?
• Are there suspicious or missing records?

Then investigate.

A practical dataset workflow

Step 1: Understand the data

Look at the columns.

Find out what each field represents.

Check data types.

Step 2: Check data quality

Look for:

• Missing values
• Duplicate records
• Impossible dates
• Incorrect categories
• Outliers
• Inconsistent spelling
• Unexpected values

Step 3: Define the business question

Don’t analyze everything just because you can.

Choose a question that matters.

Step 4: Analyze

Use SQL, Excel, Python, Power BI, Tableau, or whatever tools are relevant to the role.

Step 5: Explain the result

This is the part candidates often skip.

Don’t simply say:

“Revenue increased 14%.”

Explain:

“Revenue increased 14%, primarily because repeat purchases increased in the second half of the quarter.”

Then explain what the business might consider doing next.

That final connection between data and decision-making is extremely important.

Power BI & Tableau Interview Questions

For business intelligence roles, interviewers often want to know whether you can do more than make attractive dashboards.

A dashboard is useful only when people can understand it and make decisions from it.

Power BI questions

Be prepared to discuss:

• Power Query
• Data cleaning
• Data modeling
• Relationships
• Star schema
• DAX
• Measures vs calculated columns
• Filters
• Slicers
• Drill-through
• Row-level security
• Dashboard performance
• Data refresh

You might be asked:

“What’s the difference between a calculated column and a measure?”

A calculated column is evaluated for each row and becomes part of the model.

A measure is calculated dynamically based on the filter context when it’s used in a report.

Don’t stop there.

Explain why you’d choose one over the other.

Tableau questions

For Tableau, prepare around:

• Dimensions and measures
• Calculated fields
• Filters
• Parameters
• Dashboards
• Worksheets
• Data blending
• Relationships
• Extracts
• Dashboard performance

You may also be asked to explain why you chose a particular visualization.

That’s a surprisingly important question.

A good answer might be:

“I used a line chart because the main question was how the metric changed over time. A bar chart would make category comparison easier, but the time trend was the primary focus here.”

That shows you’re thinking about communication rather than decorating a dashboard.

Statistics & Probability Concepts

You don’t necessarily need advanced mathematics for most entry-level data analyst interviews.

But you should be comfortable with fundamental statistical concepts.

Review:

• Mean
• Median
• Mode
• Range
• Variance
• Standard deviation
• Percentiles
• Correlation
• Probability
• Distributions
• Sampling
• Confidence intervals
• Hypothesis testing
• p-values
• Regression basics
• Outliers

Mean vs median

This is a classic interview topic.

Suppose five people earn:

$30,000, $32,000, $35,000, $38,000, and $300,000.

The mean is heavily affected by the extremely high value.

The median gives you a better picture of the middle observation.

So if an interviewer asks which metric you’d use, don’t automatically say one is better.

Explain the context.

That’s the pattern you should use throughout your interview.

Don’t just define a concept. Explain when it matters.

Correlation isn’t causation

You should be comfortable explaining this without sounding like you’re reciting a textbook.

If ice cream sales increase when swimming pool attendance increases, that doesn’t necessarily mean buying ice cream causes people to go swimming.

There may be another variable involved: hot weather.

This kind of reasoning is exactly what interviewers want to see when they ask analytical questions.

Case Study Interviews – How to Solve

Case studies can feel intimidating because there often isn’t one perfect answer.

That’s actually the point.

The interviewer wants to observe your process.

Suppose you’re given this scenario:

“An online retailer’s sales have fallen by 15%. What would you investigate?”

Don’t immediately start guessing.

Break the problem down.

Step 1: Clarify the problem

Ask:

• Which products?
• Which markets?
• Compared with what period?
• Is the decline in revenue, orders, or both?
• Did traffic change?
• Did conversion rates change?

Step 2: Break the problem into categories

You could investigate:

Traffic

Did fewer people visit the website?

Conversion

Are visitors purchasing less often?

Average order value

Are customers spending less per purchase?

Product availability

Are popular products out of stock?

Pricing

Did prices change?

Competition

Did a competitor launch a promotion?

Seasonality

Is the comparison being made against an unusually strong period?

Step 3: Request or identify relevant data

Now decide what information would help answer the question.

This might include:

• Website traffic
• Conversion rate
• Orders
• Revenue
• Average order value
• Product availability
• Marketing spend
• Customer segments
• Geographic data

Step 4: Form hypotheses

Don’t randomly search for patterns.

Create possible explanations and test them.

Step 5: Communicate the recommendation

Finish with something practical.

For example:

“The largest decline appears to come from mobile conversion rather than overall traffic. I’d investigate the mobile checkout experience and compare the timing against recent website changes.”

Even if your conclusion isn’t perfect, a structured approach demonstrates analytical thinking.

Behavioral Questions for Data Roles

Technical knowledge gets you only part of the way.

You also need to prepare for questions such as:

• Tell me about yourself.
• Tell me about a difficult data problem.
• Describe a mistake you made.
• Tell me about a time you disagreed with someone.
• How do you explain technical findings to non-technical people?
• Tell me about a project where the data was incomplete.
• How do you prioritize multiple requests?
• Tell me about a time your analysis changed a decision.

Use real examples whenever possible.

The STAR method is useful:

Situation: What was happening?

Task: What were you responsible for?

Action: What did you actually do?

Result: What happened afterward?

Don’t turn every answer into a five-minute story.

A concise example with a clear result is usually stronger.

For data roles, make sure your story includes your thinking.

For example:

“The marketing team believed the campaign was performing well because clicks were high. I looked at the conversion data and noticed that traffic was increasing but purchases weren’t. I segmented the results by device and found that mobile users were converting significantly less. We investigated the mobile landing page and discovered a form issue.”

That story demonstrates analysis, communication, and business impact.

Salary Negotiation Tips (USA & UK)

Salary discussions can be uncomfortable, especially when you’re focused on getting your first data job.

But don’t approach negotiation as a confrontation.

Your first step should be research.

Look at salary information from multiple sources rather than trusting one number.

Useful sources can include:

• Glassdoor
• Indeed
• LinkedIn
• Levels.fyi
• Otta/Welcome to the Jungle where relevant
• Salary reports from recruiters
• Industry-specific salary surveys

For US roles, consider the location carefully. A data analyst salary in New York can be very different from one in a smaller market.

For UK roles, pay attention to location as well, particularly the difference between London and other regions.

Also consider the entire package.

Look at:

• Base salary
• Bonus
• Pension contributions
• Health benefits
• Equity or stock
• Paid leave
• Remote or hybrid arrangements
• Learning budget
• Flexible working
• Relocation support

What should you say?

If an employer asks:

“What are your salary expectations?”

You don’t have to immediately give one exact number.

You can say:

“Based on the responsibilities of the role, the market range I’ve researched, and my experience with SQL, Excel, and BI tools, I’m targeting a range of X to Y. I’m also interested in understanding the overall compensation package.”

The important part is replacing X to Y with numbers supported by your actual research.

Don’t choose a number because it sounds impressive.

Choose a range based on the role, location, experience level, and total compensation.

Don’t negotiate too early

If you’re still at the initial screening stage, your priority is understanding whether the job is a good fit.

Once you have an offer, your position is usually much stronger because the employer has already decided they want you.

And remember: negotiation isn’t only about demanding more money.

You might negotiate flexibility, additional leave, professional development funding, or a review after a defined period.

From Interview to First Day Success

Getting the offer is a major achievement.

But your first day can feel strangely similar to your first interview.

You don’t know where everything is.

You don’t know who everyone is.

You don’t know which systems to use.

And you probably don’t understand half the acronyms people casually throw around.

Don’t panic.

Nobody expects a new employee to understand the entire organization immediately.

Before your first day

Find out:

• Start time
• Office or remote arrangements
• Required equipment
• Login information
• Dress expectations
• Who you’ll report to
• What documents you need

If you’re working remotely, test your microphone, webcam, internet connection, and required software before your first meeting.

It sounds basic, but avoiding technical problems on day one removes unnecessary stress.

During your first week

Focus on listening.

Learn:

• What your team does
• What metrics matter
• Which reports already exist
• Where data comes from
• Who owns different datasets
• Which stakeholders you’ll work with
• How requests are prioritized

Take notes.

A simple OneNote, Notion, Google Doc, or company-approved note-taking system can be incredibly useful.

Create your own mini knowledge base.

Don’t try to impress everyone immediately

This is another mistake I’ve seen.

A new analyst joins a company and immediately starts criticizing existing dashboards.

Maybe the dashboard really is poorly designed.

But before changing it, understand why it exists.

There may be business rules, legacy systems, stakeholder requirements, or technical limitations you haven’t discovered yet.

Ask questions first.

Then make improvements when you understand the context.

A Practical Interview Preparation Routine

If your interview is two weeks away, don’t spend every day randomly watching interview videos.

Create a simple schedule.

Days 1–3

Review SQL fundamentals and practice queries.

Days 4–5

Practice Excel tasks using a real dataset.

Days 6–7

Build or review a Power BI or Tableau dashboard.

Days 8–9

Review statistics and probability.

Days 10–11

Complete one case study under time pressure.

Day 12

Practice behavioral questions.

Day 13

Research the company and review the job description again.

Day 14

Do a full mock interview.

During the mock interview, don’t just check whether your answers are correct.

Check whether you communicate clearly.

A technically correct answer that nobody can understand isn’t very useful in a business environment.

Final Thoughts

The strongest data interviews aren’t won by memorizing the most definitions.

They’re won by demonstrating that you can think with data.

Know your SQL.

Be comfortable with Excel.

Understand the BI tool listed in the job description.

Review your statistics.

Practice real datasets.

But spend just as much time learning how to explain your reasoning.

When you receive a messy dataset, don’t panic.

When you don’t know an answer, don’t bluff.

When an interviewer challenges your assumption, don’t become defensive.

Talk through your thought process.

That’s what turns a technical interview from a memory test into a conversation about how you actually work.

And once you get the offer, remember that the interview isn’t the finish line. Your first few months are where you turn those skills on your resume into real professional experience.