What is Included in Report Refresh in Power BI?

When you refresh a report in Power BI, you're doing more than just clicking a button to see the latest numbers. This triggers a specific sequence of events that pull fresh data from your sources and update the entire data model powering your visuals. This article breaks down exactly what happens during a Power BI report refresh, what gets updated, what stays the same, and how you can manage the process.

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

First, Understand the Difference: Report vs. Dataset

Before getting into the details of a refresh, it’s critical to understand that a Power BI file is made of two main parts: the Report and the Dataset.

• The Report: This is the visual layer - the charts, tables, cards, and slicers you interact with. It’s the storytelling part of your BI solution where you arrange visuals to present insights.
• The Dataset: This is the engine under the hood. It contains your imported data, the connections to your data sources, the relationships between tables, and all your DAX calculations (measures and calculated columns).

This distinction is the most important concept to grasp. A refresh action is performed on the dataset, not the report itself. When the dataset is updated with new information, the visuals in your report automatically reflect those changes.

What a Power BI Refresh Actually Does

There are a few different types of refreshes in the Power BI ecosystem, but when most people refer to a "refresh," they are talking about a data refresh. This is the core process of updating the data model for reports using Import mode, where a copy of your data is stored within Power BI.

Here's a step-by-step look at what occurs during a full data refresh:

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

1. Connection to Data Sources

Power BI first needs to establish a connection with the original sources of your data. This could be anything from a simple Excel file on OneDrive, a SQL Server database in your office, or a SaaS platform like Salesforce. It uses stored credentials to gain access. If your data source is on premises (like a local SQL database), Power BI will use an On-Premises Data Gateway to securely connect to it.

2. Power Query Transformations Are Re-Applied

Once connected, Power BI doesn't just grab the raw data. It replays every single step you defined in the Power Query Editor. This includes:

• Removing or rearranging columns
• Filtering rows (e.g., removing nulls or errors)
• Splitting or combining columns
• Unpivoting data
• Changing data types
• Running any custom M code functions

This is a crucial point: the refresh process reruns your entire transformation script on the source data. If your data source structure has changed (for example, if a column name was changed by your IT department), the refresh will fail at this stage.

3. Data is Loaded into the Data Model

After the data is freshly transformed, it’s loaded into Power BI's internal VertiPaq analysis engine. For a standard full refresh, the existing tables in your dataset are wiped and completely replaced with the newly transformed data from your sources.

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

4. Calculated Columns and Tables are Re-Computed

With the new data loaded, Power BI must now re-evaluate any calculated columns or calculated tables you created using DAX. For instance, if you have a calculated column that categorizes sales as "Large" or "Small" based on the sales amount, this logic will be applied to every single row in the refreshed sales table.

5. Relationships are Processed and the Model is Recalculated

Finally, Power BI recalibrates the data model. It uses the existing relationships to understand how tables filter one another and updates its internal query caches, getting your dataset ready for use. While DAX measures are not pre-calculated and stored like columns are, the model is now primed to calculate them extremely quickly based on the very latest data.

Once this process is complete, any visuals in the connected report will now display the refreshed data.

What IS Included in a Data Refresh (The Updated Parts)

To put it simply, here’s a checklist of everything that gets updated when you schedule or manually start a refresh.

• New Data Rows: Any new records added to the data source since the last refresh are pulled in. If 100 new entries were added to your transaction table, they will now be in your dataset.
• Updated Data in Existing Rows: Changes made to existing records are captured. If a customer's status in your CRM changed from "Active" to "Inactive," the refresh will reflect that change in the corresponding row in Power BI.
• Deleted Data Rows: If rows were deleted from the source data, they will also be removed from your Power BI dataset after the refresh (since the old table is wiped and replaced).
• Calculated Column Values: The values within calculated columns are re-computed for every row, including all the new and updated ones.
• Calculated Table Contents: Any tables generated by DAX formulas (like a dynamic Calendar table) are completely rebuilt based on the new data landscape.

What IS NOT Included in a Data Refresh (The Static Parts)

Equally important is knowing what a refresh doesn’t touch. These are elements you must change manually within Power BI Desktop and republish.

• Report Visuals and Layout: Your choice of charts, colors, font sizes, filters, and page layout are considered part of the report design. They remain untouched during a refresh. A refresh changes the data inside the visuals, not the visuals themselves.
• Power Query Steps: The refresh executes your transformation steps, but it does not change them. If you need to add a new transformation step, change a filter in Power Query, or connect to a new column, you must do it in Power BI Desktop.
• Data Model Relationships: The relationships you have defined between tables are part of the dataset's core structure. A refresh will not create, delete, or modify these relationships.
• DAX Measure Formulas: Like Power Query steps, the formulas you've written for your DAX measures remain exactly as they are. The refresh simply provides new data on which these formulas will operate in real-time.
• Manually Entered Data: If you used the "Enter Data" feature in Power BI Desktop to create a static table, that data is part of the file and is not affected by a refresh unless you edit it in the .pbix file and republish it.

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

A Note on Different Connection Modes

Not all Power BI reports rely on imported data. Your connection mode plays a big role in what "refresh" means.

• Import Mode: This is the most common mode and is the process described in detail above. You need to set up scheduled refreshes to keep the data from getting stale.
• DirectQuery: In this mode, no data is actually stored in Power BI. Your report is connected directly to the source database. When you open or interact with a visual, Power BI sends a live query to the source. The data is always current, and a "refresh" clears the report cache and compels all visuals to re-query the live database.
• Live Connection: This mode connects to an existing data model, like SQL Server Analysis Services (SSAS) or another Power BI dataset. Like DirectQuery, the data stays in the source. The source data model has its own refresh schedule, and your Power BI report simply acts as a visualization layer on top of it.

Setting Up a Scheduled Report Refresh in Power BI Service

To avoid manually refreshing your dataset every time, you can schedule it to happen automatically.

1. Get Your Credentials Set Up: In Power BI Service, navigate to your workspace, find your dataset, and click the three dots (...) before selecting Settings.
2. Update Data Source Credentials: Find the "Data source credentials" section and click Edit credentials. You'll need to securely provide your login information so that Power BI can access the data without you being present. If your data is on premises, you’ll need to configure this through a data gateway.
3. Define Your Schedule: In the Scheduled Refresh section, toggle the feature on. You can choose the refresh frequency (daily or weekly), select a time zone, and add several specific times for the refresh to run.
4. Set Failure Notifications: You can choose a button to receive an email alert if the scheduled refresh fails. This is crucial for monitoring data quality issues.

Final Thoughts

A Power BI report refresh is fundamentally about updating your dataset by going back to the original source, reapplying all your configured cleaning and transformation steps, and loading the result into your data model. This distinction between the dataset and the visual report layer is the key to understanding how to build reliable, up-to-date reports that keep your business informed.

Manually connecting sources and configuring scheduled refreshes for each one is often a necessary chore for BI tools. That’s why we built Graphed to remove this friction entirely. We connect directly to your marketing and sales platforms like Google Analytics, Shopify, and Salesforce to power live, self-updating dashboards. There are no gateways to configure or refresh schedules to set, your data is always current, so you can focus on finding insights, not managing data pipelines.