Labs · Power BI

Lab: Solve 5 Interview-Style DAX Tasks on the Sample File and Check the Answers

By Ravindra Bagale · Last updated: 8 October 2026

Suppose we are…

Suppose we are in a 45-minute live DAX round at Accenture for a Power BI analyst role. The interviewer shares the same small electronics sales model we built in Lab 11 (24 orders, 4 cities, 4 products, a DateTable for 2025–2026) and gives five tasks. The numbers are sample data made up for practice.

Goal of this lab

Write five measures and check each against an exact answer:

1. Sales in 2026 only, whatever the Year slicer says.
2. Each city's % of all-city sales.
3. Rank products by sales.
4. Year-to-date sales.
5. Sales for the same period last year and the YoY %.

What you need (all free)

• Power BI Desktop and Lab-11-star-schema.pbix from Lab 11 (pbi_sales, pbi_products, pbi_cities, a marked DateTable and the Total Sales measure). No file? Do Lab 11 first; the CSVs are here:
• Download pbi_sales.csv (24 orders) Download pbi_products.csv Download pbi_cities.csv
• 45 minutes.

The data: before and after

Before. The five tasks, in plain words, as an interviewer would say them.

After. The exact values you should get.

The formula

All five use the one base measure, Total Sales = SUM ( pbi_sales[Sales] ), plus one of these ideas:

Try writing each measure before opening its answer.

Steps

1. Open Lab-11-star-schema.pbix and Save as Lab-31-dax-tasks. Add a new page. Check that DateTable is marked as a date table (Table tools → Mark as date table).

2. Task 1. Create a measure Sales 2026 that shows 2026 sales even if a Year slicer says 2025. Put it in a Card, then add a slicer with DateTable[Year] and pick 2025.

   👁 What you should see: the card shows 484K (484,000) with no slicer selection and with 2025 selected.

   Answer 1

   Sales 2026 = CALCULATE ( [Total Sales], DateTable[Year] = 2026 )

   CALCULATE replaces the filter on DateTable[Year], so the slicer cannot change it. Year is a number (YEAR returns a number), so write 2026 without quotes.

3. Task 2. Create City % of Total. Build a Table with pbi_cities[City], Total Sales and the new measure, formatted as a percentage with 2 decimals.

   👁 What you should see: Mumbai 33.72%, Pune 26.57%, Nagpur 19.97%, Nashik 19.74%, total 100.00%.

   Answer 2

   City % of Total =
   DIVIDE ( [Total Sales], CALCULATE ( [Total Sales], ALL ( pbi_cities ) ) )

   The numerator keeps the row's city (439,000 for Mumbai). The denominator removes all city filters (1,302,000). 439,000 / 1,302,000 = 33.72%. DIVIDE returns blank instead of an error if the denominator is 0.

4. Task 3. Create Product Rank and build a Table with pbi_products[Product], Total Sales and Product Rank.

   👁 What you should see: Laptop 1 (750,000), Mobile 2 (390,000), Smartwatch 3 (90,000), Headphones 4 (72,000).

   Answer 3

   Product Rank = RANKX ( ALL ( pbi_products[Product] ), [Total Sales], , DESC, Dense )

   ALL gives RANKX the full list of 4 products to compare with. Without ALL, each row sees only itself and every product gets rank 1. The total row shows 1 too; to blank it, wrap it: IF ( HASONEVALUE ( pbi_products[Product] ), RANKX ( … ) ).

5. Task 4. Create Sales YTD. Build a Matrix: Rows = DateTable[Year] then DateTable[Month] (Month sorted by Month No, as in Lab 12), Values = Total Sales and Sales YTD. Expand 2025.

   👁 What you should see: June 2025 Sales YTD = 475,000 (60k + 120k + 20k + 150k + 95k + 30k), and December 2025 = 818,000. January 2026 starts again at 100,000.

   Answer 4

   Sales YTD = TOTALYTD ( [Total Sales], DateTable[Date] )

   TOTALYTD needs the Date column of a proper date table. It restarts each 1 January. If the matrix uses pbi_sales[OrderDate] instead of DateTable, the numbers go wrong: always slice by DateTable.

6. Task 5. Create Sales LY and YoY %. Add both to the matrix and expand 2026.

   👁 What you should see: January 2026: Total Sales 100,000, Sales LY 60,000, YoY % 66.7%.

   Answer 5

   Sales LY = CALCULATE ( [Total Sales], SAMEPERIODLASTYEAR ( DateTable[Date] ) )
   YoY % = DIVIDE ( [Total Sales] - [Sales LY], [Sales LY] )

   (100,000 − 60,000) / 60,000 = 66.7%. Look at the 2026 year row: Sales LY shows 818,000, all of 2025, because 2026 in DateTable means all 12 months. But 2026 data only runs to June, so YoY % shows −40.8%, which is misleading. That is the trap in the challenge below.

7. Save the file. Count your score: 1 point for each task you solved without opening the answer.

Common mistakes

Self-check checklist

Try-at-home challenge

The interviewer asks: "Is 2026 better or worse than 2025 so far?" The year row says −40.8%. Give a fair answer with a number: compare January–June 2026 with January–June 2025.

Samjla ka? CALCULATE, ALL, RANKX, TOTALYTD, SAMEPERIODLASTYEAR: five tasks, five exact numbers. Aata pudhe jaauya: rebuild your first bar chart with keyboard shortcuts.