Graphed LogoGraphed
Graphed LogoGraphed
How It WorksAgentsIntegrationsPricing

How to Transform Data in Power BI Web Version

Cody Schneider••8 min read

Transforming your messy raw data directly in the Power BI web version used to be a dream. Now, using Power Query Online, you can clean, shape, and remodel your data right from your browser. This tutorial walks you through how to use the built-in transformation tools inside the Power BI service to get your data report-ready without ever opening the desktop app.

GraphedGraphed

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.

Watch Graphed demo video

What is Power Query Online?

If you've used Power BI Desktop, you're likely familiar with the Power Query Editor. It's the engine room where you clean up your data before building visuals. Power Query Online is the browser-based version of this powerful tool, built directly into the Power BI service (the web version).

You access it primarily through a feature called dataflows. A dataflow is essentially a collection of data tables that you've prepared in the cloud using Power Query Online. Once you create a dataflow, you and your colleagues can connect to it as a clean, reliable data source for creating new Power BI reports, saving everyone from repeating the same cleaning steps.

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.

Getting Started: Your First Dataflow

The journey to transforming data in the web starts with creating a dataflow. This is where you'll connect to your data and launch the Power Query editor.

Here's how to do it:

  1. Log in to your Power BI account at app.powerbi.com.
  2. Select a Workspace where you have authoring permissions (you can't create them in "My Workspace"). If you don't have one, create a new one by clicking Workspaces > New workspace.
  3. From within your workspace, click + New in the top-left corner and select Dataflow.
  4. A new screen will appear. Click the Add new tables button.
  5. Next, you'll be prompted to choose a data source. There are many options, from databases to web pages. For this example, let's use a common one: an Excel workbook stored in a OneDrive or SharePoint folder. Select a source and follow the prompts to connect.

Once you connect and select the tables or sheets you want to import, Power BI will load a preview and present a button to Transform data. Click it, and you'll be taken to the Power Query Online interface, which looks very similar to its desktop counterpart.

The Power Query Online Interface Explained

When the editor opens, you'll see a few key areas that are essential to understand for transforming your data effectively.

The Main Components:

  • Ribbon: At the top of the screen, the ribbon contains all the transformation tools, organized into tabs like Home, Transform, Add Column, and View. This is where you'll find everything you need to manipulate your data.
  • Queries Pane: On the left side, this pane lists all the queries (data tables) in your dataflow. You can select different queries here to work on them.
  • Data Preview: The large central area shows a preview of your selected query's data. You can interact with columns and cells directly from this view. Notice the little icons next to the column headers - these let you filter data and change data types.
  • Query Settings Pane: This pane on the right is your co-pilot. It has a section for the table's name (under Properties) and, most importantly, a list called Applied Steps. Every single change you make is recorded here as a step, creating a repeatable recipe for your data transformation.

Common Data Transformations Step-by-Step

Let's dive into the most common data cleaning tasks you'll perform. All of these transformations add a new step to the Applied Steps pane, so you can always undo or review your work.

1. Managing Columns: Choose, Remove, and Rearrange

Often, your source data has more columns than you need. A clean dataset only includes relevant information.

  • To Remove Columns: Select the column(s) you want to remove by clicking on their headers (hold Ctrl to select multiple). Right-click on one of the headers and choose Remove Columns.
  • To Choose Columns: An easier way to remove many columns at once is to pick the ones you want to keep. Go to the Home tab on the ribbon and click Choose Columns. A dialog box will appear where you can uncheck all the columns you don't need.
  • To Reorder Columns: Simply click and drag a column header to the left or right to change its position.
  • To Rename a Column: Double-click on the column header and type in a new name, then press Enter.

2. Filtering and Removing Unwanted Rows

Dirty data often contains blanks, errors, or irrelevant entries. Filtering is essential for cleanliness and accuracy.

  • Remove Blank Rows: Go to the Home tab and click on the Remove Rows dropdown. Select Remove Blank Rows. This will automatically find and delete any row where all cells are empty.
  • Filter by a Value: Click the small arrow icon on the column header you want to filter. A dropdown will appear, similar to Excel, where you can uncheck values you wish to exclude. You can also use the Text Filters or Number Filters options for more advanced conditions like "Does Not Contain," "Begins With," or "Is Greater Than."
GraphedGraphed

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.

Watch Graphed demo video

3. Changing Data Types

Power BI needs to understand what kind of data is in each column to perform calculations and create charts correctly. Dates should be dates, numbers should be numbers. This is one of the most critical steps.

To change a data type:

  1. Look at the icon to the left of the column name in the header (e.g., ABC for text, 123 for whole number, a calendar for date).
  2. Click on that icon.
  3. A menu will appear with a list of data types. Select the correct one for that column, such as Date, Decimal Number, or True/False.

Getting this right prevents countless reporting headaches later on.

4. Text Transformations

Text data often comes in inconsistent formats. Power Query has excellent tools to fix this in the Transform and Add Column tabs.

Example: Splitting a 'Full Name' Column

Imagine you have a column with "First Last" and you need separate First Name and Last Name columns.

  1. Select the "Full Name" column.
  2. Go to the Transform tab ribbon and click Split Column > By Delimiter.
  3. Power Query is smart and will often automatically detect the space as the delimiter. Confirm an option like "Each occurrence of the delimiter" and click OK.
  4. Now you have two new columns, which you can rename to "First Name" and "Last Name."

Other Useful Text Transformations:

  • Format: In the Transform tab, use the Format button to quickly change text case (lowercase, UPPERCASE, Capitalize Each Word).
  • Trim & Clean: Tucked inside the Format menu, Trim removes extra spaces from the beginning and end of a text string, while Clean removes non-printable characters. These are lifesavers for fixing hidden formatting issues.
  • Merge Columns: Need to do the opposite of splitting? Select two or more columns, right-click, and choose Merge Columns. You can specify a separator to place between the values being combined.

5. Working with Columns from Examples

This is where Power Query feels like magic. Instead of figuring out the right buttons to click, you can just show Power Query what you want an outcome to look like, and it will figure out the transformation for you.

Example: Create a Full Name from First and Last Name

  1. Go to the Add Column tab on the ribbon.
  2. Click Column from Examples.
  3. A new, blank column will appear on the right.
  4. In the first row of this new column, start typing what you want the final string to look like. For example, if "First Name" is "John" and "Last Name" is "Smith", type "John Smith".
  5. As you press Enter and type "Jane Doe" for the second row, Power Query will likely recognize the pattern and automatically fill in the rest of the column for you. The formula it generated appears at the top.
  6. If it looks correct, click OK. You've just created a merged column without writing any code.

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.

Saving Your Work

Once you've applied all your desired transformations, click the Save & close button in the bottom right corner of the screen. Power BI will then process your steps, and the transformed table will be saved in your dataflow.

You can set your dataflow to refresh on a schedule, ensuring your reports are always built from a clean, up-to-date data source without any manual data prep on your part.

Final Thoughts

Mastering data transformation in Power BI's web version liberates you from the desktop application for many common data prep tasks. By using dataflows and Power Query Online, you can create a reusable, centralized, and automated process for cleaning data, which is a major step toward building reliable and refreshable reports for your team.

While learning tools like Power BI unlock powerful analytics capabilities, sometimes you need immediate answers without getting stuck in the weeds of transformation steps. We built Graphed for this exact reason. You can securely connect your data sources like Google Analytics, Shopify, or Salesforce, and our AI does the heavy lifting. Instead of building dataflows, you just ask questions in plain English - like "what were our top 5 campaigns by ROI last month?" - and get an interactive dashboard instantly, letting you focus on the insights, not the setup.

Related Articles