How to Validate Data with Group By in Power BI

One of my students recently asked me: “How can we validate data in Power BI?” It’s an excellent question, and one of the simplest answers is to validate data with Group By in Power BI. This approach allows you to check totals, confirm consistency, and quickly spot problems in your data model. Along with Group By, you can also use Column Quality in Power Query to quickly spot errors, blanks, and unexpected values.

Steps to Validate Data with Group By in Power BI (Quick Guide)

1. Load the table into Power Query
2. Create a copy (reference) of the table
3. Apply Group By on different dimensions (e.g. Orders, Products)
4. Create visualizations in Power BI Desktop and compare totals to validate results

New to Power BI? Get hands-on, step-by-step guidance in our Power BI for Beginners course. Start learning now.

Why Validate Data in Power BI?

When building reports, errors often hide in the data model. Common issues include:

• Duplicate rows in dimension tables
• Missing or orphan keys between tables
• Incorrect granularity in fact tables
• Inflated totals from many-to-many relationships

By validating early, you can fix these problems before they affect your analysis.

👉 Related reading: How to Create Relationships in Power BI

What is Group By in Power BI?

In Power Query, Group By lets you summarize data, much like the GROUP BY clause in SQL. You can roll up values by one or more columns and calculate sums, counts, or averages.

This makes Group By an excellent tool not just for transformation, but also for validation. By grouping your data across different dimensions and comparing totals, you can quickly detect errors.

👉 If you’re new to Power Query, start with our guide: Loading Data Using Query Editor in Power BI

How to Validate Data with Group By in Power BI (Step-by-Step)

1. Loading the Table in Power BI. For this exercise, I will load OrderDetails table using Query Editor. This table contains around 102 rows of data and has the following structure:

2. Create a copy of the table. Once the table is loaded in Query Editor create a copy of it by using reference. Right-click on OrderDetails and select Reference as shown in the figure below:

Rename this Query as TotalSales_ByOrderID

3. Group the table by OrderID. With TotalSales_ByOrderID query selected, navigate to the ribbon and select Transform/Group By. In the Group By dialog box, make selections as shown in the figure below:

This query is similar to database query:

Select OrderID, Sum(Sales) FROM OrderDetails Group By OrderID;

4. Create another copy of OrderDetails by right-clicking on it and selecting Reference. Rename this Query as TotalSales_ByProductID.

5. Group the table by ProductID. With TotalSales_ByProductID query selected, navigate to the ribbon and select Transform/Group By. In the Group By dialog box, make selections as shown in the figure below:

This query is similar to database query:

Select ProductID, Sum(Sales) FROM OrderDetails Group By ProductID;

6. Create another copy of OrderDetails by right-clicking on it and selecting Reference. Rename this Query as TotalSales_ByOrderID_ProductID.

7. Group the table by OrderID and ProductID. With TotalSales_ByOrderID_ProductID query selected, navigate to the ribbon and select Transform/Group By. In the Group By dialog box, make selections as shown in the figure below:

This query is similar to database query: Select OrderID, ProductID, Sum(Sales) FROM OrderDetails Group By OrderID, ProductID;

8. To apply your changes in Query Editor, navigate to the ribbon and select File/Close & Apply. This action will navigate you back to Power BI desktop.

9. Creating Visuals to validate the data. In the Power BI desktop, you will notice 3 more tables apart from OrderDetails. These tables will be TotalSales_ByOrderID, TotalSales_ProductId, and TotalSales_OrderID_ProductID.

Create table visualizations using these table and check the Total as shown in the figure below:

Notice in the visuals above, it shows same Totals by different groupings.

Interpretation of the results

We grouped OrderDetails three different ways:

• By OrderID → Total Sales by each order.
• By ProductID → Total Sales by each product.
• By OrderID + ProductID → Total Sales by order-product line items.

In all cases, the grand total Sales was the same.

This shows that:

• OrderDetails table has no duplicate rows beyond the expected grain (OrderID + ProductID).
• Sales column is consistent (no hidden extra or missing values).
• Aggregations are reliable no matter which dimension you roll up by.
• The fact table is clean and well-grained (no accidental many-to-many issues inside OrderDetails).

Interpretation if the totals were different

If totals had differed values, it would signal:

• Duplicate dimension keys (double-counting when grouped differently).
• Granularity issues (extra detail rows causing inflated totals).
• Bad relationships (if done via joins).

But since totals are consistent → shows fact table is stable, and Power BI will give consistent numbers across visuals.

Validating data with Group By in Power BI ensures totals stay consistent, the fact table has the right grain, and the model delivers reliable results. Matching totals across Orders, Products, and Order–Product groupings confirm data quality, while mismatches highlight duplicates, grain issues, or relationship errors.

Group By is one way to validate data, but it’s not the only option. Another useful method is available in Power Query. Learn more in our guide on How to Validate Data in Power BI with Column Quality

👉 Want to practice this yourself? Create your own data, or let us know if you’d like the OrderDetails.xlsx sample dataset, and follow the steps above.
👉 Ready to take your Power BI skills further? Enroll in our Power BI for Beginners course and start building accurate, professional dashboards today.

---