How to Edit Dataset in Power BI Service

You've done the hard work. You connected your data, built your model in Power BI Desktop, created beautiful visuals, and finally published your masterpiece to the Power BI Service for your team to use. Then, you get a message: "Can we rename this column?" or "Could you add a quick calculation for profit margin?" Your first thought might be to sigh, open the original .pbix file, make the change, and re-publish. But you don't always have to do that.

Graphed

Build AI Agents for Marketing

Build virtual employees that run your go to market. Connect your data sources, deploy autonomous agents, and grow your company.

Build Your First Agent

You can actually make several important edits to your dataset directly from your web browser in the Power BI Service. This guide will walk you through exactly what you can change, what you can’t, and how to make those quick fixes without ever leaving the cloud.

Understanding the Two Power BI Worlds: Desktop vs. Service

Before we jump into the "how," it's crucial to understand the fundamental difference between Power BI Desktop and the Power BI Service. Thinking of them as separate tools for different stages of the analytics process will save you a lot of headaches.

Power BI Desktop is your workshop. It’s a free, downloadable application for your computer where you do the heavy lifting. This is where you:

• Connect to raw data sources (Excel files, SQL databases, web APIs, etc.).
• Transform and clean that data using the Power Query Editor (ETL processes like splitting columns, removing errors, unpivoting data).
• Build your data model, define relationships between tables, and write complex DAX measures.
• Design the layout of your reports and visuals.

Power BI Service is your showroom and collaboration hub. It's the cloud-based platform where you publish, share, and consume reports. This is where you:

• Create and organize workspaces for different teams or projects.
• Share reports and dashboards with colleagues and stakeholders.
• Set security and manage user permissions.
• Schedule automatic data refreshes to keep your reports up to date.
• Perform light-touch editing on datasets and reports.

The key takeaway is that Power BI Service is primarily for sharing and consumption, while a majority of the development happens in Desktop. However, Microsoft has been adding more and more editing capabilities to the Service, and knowing how to use them can be a massive time-saver for common reporting adjustments.

AI Agents for Marketing

Crash Course

Graphed

Free PDF · the crash course

AI Agents for Marketing Crash Course

Learn how to deploy AI marketing agents across your go-to-market — the best tools, prompts, and workflows to turn your data into autonomous execution without writing code.

Download the PDF

Editing Your Dataset in the Power BI Service: The How-To Guide

So, what exactly can you edit once a dataset is live in the Power BI Service? Quite a lot, actually. You can create calculations, tweak your data model, and update formatting right from your browser.

To access these features, navigate to your workspace in the Power BI Service, find the dataset you want to edit (it will have an orange icon), click the three-dot menu (...) next to it, and select "Open data model."

This will open a new browser tab with a familiar interface that looks a lot like the Model and Data views from Power BI Desktop. From here, you can make the following changes.

1. Create New Measures with DAX

One of the most powerful features available online is the ability to write new DAX measures. Did you forget to add a "Total Cost" calculation or need a quick "Year-over-Year Growth" percentage? You don't need to re-publish.

How to do it:

1. With the data model open, make sure you're in the "Data" pane on the right-hand side.
2. Select the table you want the measure to live in.
3. Click the "New measure" button in the ribbon at the top of the page.
4. The DAX formula bar will appear. Write your formula and press Enter.

For example, to create a simple "Total Revenue" measure in your Sales table, you’d write:

Total Revenue = SUM(Sales[RevenueAmount])

The new measure will appear in your fields list, ready to be used in any report connected to this dataset. It's a game-changer for responding to quick analytical requests without breaking your workflow.

Graphed

Build AI Agents for Marketing

Build virtual employees that run your go to market. Connect your data sources, deploy autonomous agents, and grow your company.

Build Your First Agent

2. Create New Calculated Columns with DAX

Like measures, you can also add new calculated columns to your tables. Remember the difference: a calculated column computes a value for each individual row in your table and is saved within the model, while a measure computes a value in aggregate based on the context of the visual it's in (like filters or slicers).

How to do it:

1. Select the table where you want to add the new column.
2. Click the "New column" button from the top ribbon.
3. Write your DAX formula in the formula bar.

For instance, if your Sales table has RevenueAmount and ProductCost columns, you could create a Profit column:

Profit = Sales[RevenueAmount] - Sales[ProductCost]

This will calculate the profit for every single sale on a row-by-row basis.

3. Manage and Create Table Relationships

Often, you might need to adjust or create a relationship between tables after publishing. Maybe you brought in a new lookup table or realized you connected two tables on the wrong columns. You can fix this directly in the Service's data model view.

How to do it:

1. In the top-left corner of the data model editor, switch from Table view to Model view (the icon with connecting lines).
2. You'll see a diagram of all your tables and their relationships.
3. To create a new relationship, simply drag a field from one table and drop it onto the corresponding field in another table.
4. To edit an existing relationship, right-click the line connecting the tables and choose "Properties." A pane will open where you can adjust things like cardinality (one-to-one, one-to-many) and cross-filter direction.

This is incredibly useful for refining how your tables interact without going back to Desktop.

4. Rename Tables, Columns, and Measures

Clarity is everything in a report. Field names like sum_revenue_usd or tbl_dim_prod aren't user-friendly. In the Service, you can easily rename almost anything in your data model.

• To rename a table or a column: In the Data pane on the right, find the item you want to rename, click the three-dot menu, select "Rename," and type the new name.
• To rename a measure: You simply edit its name in the DAX formula bar by changing the text before the equals sign.

These changes will instantly cascade through to all connected reports, helping you improve readability on the fly.

5. Update Formatting and Data Types

Clean formatting makes data understandable. A raw number like 45000.75 is just a number, but $45,000.75 is clearly money. You can control these display settings online.

How to do it:

1. Select the column or measure you want to format from the field list.
2. A contextual "Tools" ribbon will appear at the top.
3. From here, you can change the Data Type (e.g., Whole Number, Text, Date), Format (e.g., Currency, Percentage), and add custom formatting like thousand separators or decimal places.

This quick tweak can dramatically improve the professional look of your reports.

AI Agents for Marketing

Crash Course

Graphed

Free PDF · the crash course

AI Agents for Marketing Crash Course

Learn how to deploy AI marketing agents across your go-to-market — the best tools, prompts, and workflows to turn your data into autonomous execution without writing code.

Download the PDF

The Red Line: What You Still Can't Do in Power BI Service

While the online editing capabilities are powerful, they aren't limitless. The "workshop" of Power BI Desktop is still necessary for the heavy-duty data preparation tasks. Here are the things you cannot do in the Power BI Service and must do in Desktop:

• Use the Power Query Editor: Power Query is the heart of Power BI's data transformation capabilities. Any step that involves shaping the raw data—like splitting columns, filtering rows before they're loaded, cleaning messy text, pivoting/unpivoting, or merging multiple data files—has to be done in Power BI Desktop. The Power Query Editor does not exist in the Service.
• Change or Add New Data Sources: Need to switch your data source from a local CSV file to a live SQL Server database? Or add a new data source? That's a fundamental change to the model's foundation, and you must do it in the original .pbix file.
• Create or Manage Security Roles (RLS): Setting up Row-Level Security to restrict what certain users can see requires Power BI Desktop. While you assign users to those roles in the Service, the roles themselves have to be defined in the Desktop file first.
• View Data Lineage Outside of a Workspace: Desktop offers a more detailed data model view and dependency traceability, which helps when debugging complex models.

The Desktop-to-Service Workflow

When you encounter a task that requires one of the above changes, the workflow remains the same:

1. Open the original .pbix file in Power BI Desktop.
2. Perform your edits (e.g., make changes in Power Query).
3. Click "Close & Apply" to load the changes into your data model.
4. Click "Publish" from the Home ribbon.
5. Select the same workspace you published to before. Power BI will ask if you want to replace the existing dataset and reports. Choose "Replace."

This overwrites the cloud version with your updated local version, bringing your big changes online.

Final Thoughts

Knowing where to make your edits is key to a smooth and efficient Power BI workflow. For quick calculations, relationship tweaks, and formatting fixes, the Power BI Service data model editor is a fantastic time-saver. It keeps you in the browser and updates reports instantly. For any heavy lifting related to data sources or transformation, Power BI Desktop remains your essential tool.

Feeling the friction that comes from using several disconnected tools is common in data analytics. Many reporting workflows involve downloading CSV files from one platform, cleaning them in Excel, building a model in Power BI Desktop, and finally publishing it, only to repeat the process every week. At Graphed, we built a solution to remove that entire manual loop. Graphed connects directly to marketing and sales sources like Salesforce, HubSpot, and Google Analytics, allowing you to create real-time, self-updating dashboards by just describing what you want to see in plain English. No more jumping between tools, and no more needing to be a BI expert to get straight to the insights.