Why dataflows centralize data preparation in Power BI - Business Case Study

Scenario Mode

👤 Your Role: You are a business intelligence analyst at a retail company.

📋 Request: Your manager wants you to explain and demonstrate how dataflows can centralize data preparation to improve reporting efficiency and consistency.

📊 Data: You have sales data from multiple stores in Excel files and customer data from a CRM system. These data sources need cleaning and combining before analysis.

🎯 Deliverable: Create a Power BI dataflow that cleans and combines these data sources centrally, then build a simple report using the dataflow to show monthly sales by store.

Sample Data

StoreID	StoreName	SaleDate	SalesAmount
101	North Store	2024-01-05	1500
102	East Store	2024-01-06	1200
101	North Store	2024-02-10	1800
103	West Store	2024-02-15	1100
102	East Store	2024-03-01	1300
103	West Store	2024-03-05	1400
101	North Store	2024-03-10	1600
104	South Store	2024-03-12	1700

1

Step 1: Create a new dataflow in Power BI service.

Go to Power BI workspace > New > Dataflow > Define new entities.

Expected Result

A blank dataflow ready to add data entities.

2

Step 2: Connect to the Excel sales data files and import the sales table.

Select 'Add new entities' > Choose Excel files from OneDrive or SharePoint > Select sales table.

Expected Result

Sales data loaded into the dataflow as an entity.

3

Step 3: Clean the sales data by removing any rows with missing StoreID or SalesAmount.

In Power Query editor, filter out rows where StoreID or SalesAmount is null or blank.

Expected Result

Sales data contains only complete rows with StoreID and SalesAmount.

4

Step 4: Connect to the CRM system to import customer data as another entity.

Add new entity > Connect to CRM data source > Select customer table.

Expected Result

Customer data loaded into the dataflow as a separate entity.

5

Step 5: Create a relationship between sales and customer data using StoreID.

In dataflow, define StoreID as key to link sales and customer entities.

Expected Result

Dataflow entities linked for combined analysis.

6

Step 6: Save and refresh the dataflow to apply changes and prepare data centrally.

Click Save > Refresh dataflow to load cleaned and combined data.

Expected Result

Dataflow contains clean, combined data ready for reporting.

7

Step 7: In Power BI Desktop, connect to the dataflow as a data source.

Get Data > Power BI dataflows > Select the created dataflow.

Expected Result

Dataflow data loaded into Power BI Desktop.

8

Step 8: Create a report visual showing monthly sales by store using the dataflow data.

Create a line chart: Axis = SaleDate (month), Legend = StoreName, Values = SUM(SalesAmount).

Expected Result

Report shows monthly sales trends for each store.

Final Result

Monthly Sales by Store

StoreName  | Jan | Feb | Mar
-----------|-----|-----|-----
North Store|1500 |1800 |1600 
East Store |1200 |     |1300 
West Store |     |1100 |1400 
South Store|     |     |1700 

✓Dataflows centralize data cleaning and combining steps, so reports use consistent data.

✓Monthly sales trends by store are easy to update by refreshing the dataflow once.

✓Central data preparation saves time and reduces errors across reports.

Bonus Challenge

Extend the dataflow to include product category data and create a report showing sales by category and store.

Show Hint

Add a new entity for product categories, link it to sales by product ID, then update the report visuals.

Practice

1. Why do Power BI dataflows centralize data preparation?
Choose the best reason.

easy

A. To replace Power BI datasets completely

B. To create visualizations faster in Power BI Desktop

C. To store data only on local computers

D. To clean and prepare data once for reuse in many reports

Solution

1. Step 1: Understand dataflows purpose

   Dataflows are designed to clean and prepare data centrally.

2. Step 2: Identify benefits of centralization

   This allows reuse of the same cleaned data in many reports, saving time and ensuring consistency.

3. Final Answer:

   To clean and prepare data once for reuse in many reports -> Option D

4. Quick Check:

   Centralizing data prep = reuse + consistency [OK]

Hint: Dataflows prepare data once for many reports [OK]

Common Mistakes:

• Thinking dataflows only speed up visualization
• Believing dataflows store data locally
• Confusing dataflows with datasets

2. Which syntax correctly creates a dataflow in Power BI service?
Choose the correct step.

easy

A. Create a report > Save as Dataflow

B. Open Power BI Desktop > File > New Dataflow

C. Go to workspace > New > Dataflow > Define entities

D. Use Power Query Editor > Export as Dataflow

Solution

1. Step 1: Identify where dataflows are created

   Dataflows are created in Power BI service workspaces, not in Desktop.

2. Step 2: Confirm correct creation steps

   The correct path is workspace > New > Dataflow > Define entities to set up dataflows.

3. Final Answer:

   Go to workspace > New > Dataflow > Define entities -> Option C

4. Quick Check:

   Dataflows created in service workspace [OK]

Hint: Create dataflows in Power BI service workspace [OK]

Common Mistakes:

• Trying to create dataflows in Power BI Desktop
• Confusing dataflows with reports or datasets
• Assuming export options create dataflows

3. Given a dataflow that cleans sales data and is used in two reports, what is the main benefit?
Choose the best output.

medium

A. Both reports use consistent cleaned data without repeating steps

B. Each report must clean data separately for accuracy

C. Dataflows slow down report refresh times

D. Reports cannot share dataflows

Solution

1. Step 1: Understand dataflow reuse

   Dataflows allow multiple reports to use the same cleaned data source.

2. Step 2: Identify benefit of reuse

   This means reports share consistent data and avoid duplicate cleaning work.

3. Final Answer:

   Both reports use consistent cleaned data without repeating steps -> Option A

4. Quick Check:

   Reuse cleaned data = consistency + efficiency [OK]

Hint: Dataflows share cleaned data across reports [OK]

Common Mistakes:

• Thinking each report cleans data separately
• Believing dataflows slow refresh
• Assuming reports cannot share dataflows

4. You created a dataflow but reports show outdated data. What is the likely issue?
Choose the best fix.

medium

A. Reports do not support dataflows

B. Dataflow refresh is not scheduled or failed

C. Dataflow entities are deleted

D. Power BI Desktop needs update

Solution

1. Step 1: Check dataflow refresh status

   Dataflows must be refreshed to update data; if not scheduled or failed, data is stale.

2. Step 2: Confirm refresh scheduling

   Scheduling or manually refreshing the dataflow fixes outdated data issues.

3. Final Answer:

   Dataflow refresh is not scheduled or failed -> Option B

4. Quick Check:

   Refresh dataflow to update data [OK]

Hint: Check dataflow refresh schedule first [OK]

Common Mistakes:

• Assuming reports don't support dataflows
• Ignoring refresh schedules
• Blaming Power BI Desktop version

5. You want to centralize customer data cleaning for multiple teams. Which approach best uses dataflows?
Choose the best solution.

hard

A. Create one dataflow with cleaned customer data, then connect all team reports to it

B. Each team creates its own dataflow with separate cleaning steps

C. Use Excel files shared by teams instead of dataflows

D. Build all cleaning inside each report's Power Query

Solution

1. Step 1: Identify centralization goal

   Centralizing cleaning means one place to prepare data for all teams.

2. Step 2: Choose best dataflow use

   Creating one dataflow with cleaned data and connecting all reports ensures consistency and saves time.

3. Final Answer:

   Create one dataflow with cleaned customer data, then connect all team reports to it -> Option A

4. Quick Check:

   One dataflow for all teams = centralization [OK]

Hint: One shared dataflow centralizes cleaning [OK]

Common Mistakes:

• Making separate dataflows per team duplicates work
• Using Excel files loses automation benefits
• Cleaning inside reports repeats effort