BI Developer Interview Questions UK: 8 Real Questions and What Interviewers Want

Before your BI Developer interview, it helps to see the questions from the interviewer's side. We have picked eight from our set of 16 and added the reason each one gets asked.

Eight BI Developer interview questions, and what the interviewer is thinking

1. How do you design a data model for a Power BI report, and why is a star schema recommended?

Why they ask: Data modelling is the foundation of every BI developer's work.

2. What is DAX filter context versus row context, and how do functions like CALCULATE, FILTER and ALL work?

Why they ask: DAX fluency is the clearest test of a BI developer's depth.

3. What is the difference between Import, DirectQuery and Direct Lake or composite models in Power BI, and how do you choose?

Why they ask: Storage mode decisions drive performance, freshness and cost.

4. Explain how you would use Power Query and SQL to prepare data, and what is query folding?

Why they ask: They want to see data preparation skill and an understanding of performance.

5. How do you tune a slow SQL query or a slow dataset refresh, and what tools do you use?

Why they ask: Performance tuning separates mid-level developers from beginners.

6. A report that has worked for months suddenly shows blank visuals and the refresh fails the night before a board meeting. What do you do?

Why they ask: They are testing calm, systematic troubleshooting under pressure.

7. Tell me about a time you built a report or data model that significantly improved how a team worked.

Why they ask: They want evidence that your BI work creates real business value.

8. What does a typical week look like for a BI developer, and what makes someone excellent in the role?

Why they ask: They want to check realistic expectations of the role.

A sample answer to question 1

How do you design a data model for a Power BI report, and why is a star schema recommended?

I start by identifying the business process and the questions to be answered, then separate the data into fact tables, which hold measurable events such as sales or tickets, and dimension tables that describe them, such as date, customer, product and location. Relationships run from dimensions to facts, usually one-to-many with single-direction filtering. I avoid many-to-many relationships and bi-directional filters unless there's a strong reason, as they can cause ambiguity and slow performance. A star schema suits Power BI because the VertiPaq engine compresses and scans narrow, denormalised dimensions efficiently, DAX filter context flows predictably, and users find the model easier to understand. I hide technical keys, use a proper date table marked as such, set sensible data types and summarisation, and use clear business names and folders for measures. I'd flatten snowflaked dimensions in Power Query or the warehouse, and keep transformation logic upstream wherever possible, so the model stays lean and reusable.

Common mistake to avoid: Importing one wide flat table; Using bi-directional filters everywhere

Want the rest?

Full model answers, key points and common mistakes for all 16 questions are on the BI Developer interview questions page. The first five are free, no sign-up needed.