Power QueryExercises

Reporting and operations analyst interview prep

Recurring reports from operational systems: reliability, data quality and turnaround.

Free questions

  1. technical

    How would you combine twelve monthly CSV exports into one table?

    A strong answer covers

    • Point a query at the folder, filter to the right extension and name pattern (skipping stray or temporary ~$ files), and parse each file's Content with the same steps.
    • Keep the file name (Source.Name) so rows can be traced.
    • Expand the parsed tables, then set types once on the combined result.
    • New months are picked up automatically on refresh.

    Avoid

    • Importing each file separately and appending by hand
    • Losing the file name
  2. data quality

    What checks would you build into a recurring report so problems surface early?

    A strong answer covers

    • Row-count and total checks compared with the previous run or the source system.
    • Anti joins to find keys missing from master data.
    • Keep Errors views or flags for values that failed conversion.
    • Schema checks: fail loudly if a required column is missing rather than silently producing blanks.

    Avoid

    • Relying on someone noticing a wrong chart
    • Suppressing all errors
  3. debugging

    Totals in your report are higher than the source system. Where do you look first?

    A strong answer covers

    • Duplicates: a merge with non-unique keys, or the same file combined twice.
    • Filters: cancelled or test records not excluded.
    • Subtotal or grand-total rows imported from a report-style export.
    • Compare row counts step by step to find where the inflation starts.

    Avoid

    • Adjusting numbers manually
    • Assuming the source system is wrong without checking
  4. prioritisation

    Three teams want report changes this week. How do you prioritise?

    A strong answer covers

    • Clarify impact and deadline for each: who is blocked, what decision depends on it.
    • Separate quick fixes from changes that need data owners.
    • Agree priorities with stakeholders openly and communicate timelines.
    • Protect correctness: don't ship a fast change that breaks existing reports.

    Avoid

    • First come, first served
    • Saying yes to everything
  5. technical

    Explain Fill Down and when you'd use it.

    A strong answer covers

    • Fill Down replaces nulls with the nearest value above.
    • Use it for report-style exports where a group label is printed once per block.
    • Apply it before sorting or filtering so rows are still in report order.
    • It fills null only, not empty text.

    Avoid

    • Using it after sorting
    • Confusing it with Replace Values

Loading more questions…