Power BI Interview Questions
Here is a list of common Power BI interview questions that are open ended and use to demonstrate the overall knowledge of the applicant. These questions are real and allow you to find intermediate level Power BI practitioners. Answering 6 out of the 10 question would demonstrate a confident familiarity with Power BI.
1. Where do you think Power BI should be used the most? analysis, visualization or data preparation?
Sample Answer:: The beauty of Power BI is that it allows you to prep and analyze your data. I use it to build complex data models by creating relationships between various imported tables.
These tables can be preprocessed/cleansed using the edit queries options and some M techniques. This preprocessing occurs in repeatable steps which can be applied each time data is refreshed.
Analysis can be done through the process of producing insightful visuals and using DAX.
2. Describe a time when Power BI limited or enhanced your success on a project?
Sample Answer: I think Power BI has issues with exporting. There is a limitation to export data created due to sampling. This makes it very difficult to export large datasets created in Power BI. I would like to build datasets based on multiple sources for export in Power BI. So, unfortunately, this is not possible so this action still has to be done server-side or through use with Python.
3. What are the advantages of calculated columns vs calculated measures?
Sample Answer: Although most things can be achieved with either. Taking in consideration the performance of the report or model, a calculated measure does not require recalculation when the model is refreshed only when the measure is used as opposed to a calculated column which take-up space by adding to the model.
It’s important to keep table tall and skinny for improved performance. Calculated columns will allow you to filter the data model if that is your intentions. Also, for row interactions calculated columns aren't needed due to the same being accomplished with iterator functions like SUMX which will perform row by row results.
4. What are your go-to DAX functions for Time Intelligence?
Sample Answer: DATEDADD is the most versatile function for time intelligence. However, I often you same time previous year. However, pretty much most of the time intelligence can be recreated with DATEADD
5. What's the difference between SUM and SUMX?
Sample Answer: SUM is used to aggregate the total of a single column. SUMX is used for an expression or calculation that may involve multiple columns. It goes row by row and calculates the sum of the result. These are known as iterators.
6. Do you have a philosophy when creating dashboards?
Sample Answer: Ensure visual represents the relationship between the variable. Maximize white space for readability, the main purpose of the dashboard should be discernable in 5 seconds of viewing. No more than 5 visuals per page.
7. How do you improve the overall performance of your data models?
Sample Answer: Following the tidy data philosophy, column usage should be limited. Also utilizing start schema to limit joins will also improve the model. Disable unused tables from loading on refresh also helps. Use measure as opposed to calculated columns unless. Also using lookup tables can improve overall functionality.
8. If you had a wishlist of changes in Power BI, what would change and why?
Sample Answer: Better exporting capabilities such as being able to export without sampling. Additionally, a preview mode in the desktop would be great. A better integration with Power BI where reports can be interactive in PowerPoint or vice versa. Map figures are limited with only 30,000 points so I would like this upgraded also.
9. Why would you use a date table?
Sample Answer: The inbuilt time intelligence requires a date table for time series comparisons. You need contiguous dates for this to work. Some data does not meet these standards.
Please feel free to add more questions in the comments that you think would be relevant.