How to Use Power BI Copilot

Cody Schneider••8 min read

Power BI’s Copilot is transforming the way we create reports and dashboards, converting complex requests into live visuals through simple conversational language. Instead of spending hours clicking, dragging, and writing formulas, you can now describe what you need, and Copilot builds it for you. This guide will walk you through exactly how to set up and use Power BI Copilot to generate reports, write DAX measures, and find insights faster.

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 Exactly is Power BI Copilot?

Think of Copilot as an AI data analyst built directly into the Power BI interface. It’s designed to be your assistant, helping you move from raw data to finished reports with less friction. Instead of manually creating visuals, formatting pages, and struggling with complex calculations, you can simply tell Copilot what you want in plain English.

Its core capabilities fall into three main areas:

  • Automated Report Creation: Give Copilot a prompt like, "Create a report showing our key sales metrics like revenue, profit, and units sold over time by product category," and it will generate a complete report page with relevant charts and KPIs.
  • DAX Simplification: Writing Data Analysis Expressions (DAX) is often the steepest part of the Power BI learning curve. Copilot can write DAX measures for you. Just ask, "Create a DAX measure for year-over-year revenue growth," and it will generate the code.
  • Data Summarization: After building a report page, you can ask Copilot to summarize the insights from the visuals. It provides a narrative summary of key trends and takeaways, which is perfect for adding context to your dashboards or prepping for a presentation.

Ultimately, Copilot is designed to democratize data analysis. It lowers the technical barrier, allowing marketers, sales managers, and business owners to build their own reports without needing to become Power BI experts. For seasoned analysts, it automates the tedious parts of report building, freeing up more time for strategic thinking and deep-dive analysis.

Getting Started: Enabling Power BI Copilot

Before you can start interacting with your data, you and your administrator need to enable Copilot. The feature is being rolled out and has some specific requirements, so it might not be available to everyone just yet.

Requirements and Prerequisites

First, ensure your organization meets the technical requirements. Copilot functionality depends on your Microsoft Fabric capacity.

  • Your organization must have a Power BI Pro license.
  • Copilot requires your workspace to be on a Fabric Capacity (F64 or higher) or a Power BI Premium Capacity (P1 or higher).
  • The feature is currently available for workspaces in specific regions, which are expanding over time.

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.

Steps for Admins to Enable Copilot

If you're a Power BI or Fabric administrator, here’s how you turn on Copilot for your organization:

  1. Navigate to your Fabric Admin portal.
  2. Go to the "Tenant settings" and find the section for "Copilot and Azure OpenAI Service settings."
  3. Enable the main toggle for "Copilot." You will likely need to also enable the setting that allows data to be processed by Azure OpenAI services outside your geographic region, as Fabric capacities are not yet available everywhere.
  4. You can choose to enable it for the entire organization or for specific security groups. It's often a good practice to roll it out to a smaller group of power users first.

Once the admin has enabled it, individual report creators can then activate it within their projects.

Activating Copilot in Your Report

With tenant-level settings enabled, you can now toggle Copilot on in Power BI Desktop or the Power BI service.

  1. Open a report in Power BI Desktop.
  2. Go to File > Options and settings > Options.
  3. Under the "Global" settings, select Preview features.
  4. Check the box next to "Copilot". You'll need to restart Power BI Desktop for the change to take effect.
  5. Once restarted, you should see the Copilot button in the ribbon. Click it to open the Copilot pane and begin.

Creating Report Pages with Natural Language Prompts

This is where the magic happens. A blank report page is no longer intimidating, it’s an open conversation waiting to happen. The Copilot pane acts as your command center for report generation.

Starting Your First Prompt

Imagine you're a marketing manager looking at advertising campaign data. You want a high-level overview of performance.

With an empty report page, open the Copilot pane. You can start with a broad request. For example:

"Create a page to analyze our marketing campaign performance. Include total investment, total impressions, and total conversions."

Copilot will analyze this prompt, identify the relevant fields in your data model (like 'Investment', 'Impressions', 'Conversions'), and generate a report page. It thoughtfully selects what it thinks are the best visuals - perhaps a few KPI cards for the totals, a bar chart showing investment by campaign, and a line chart tracking conversions over time.

It’s important to remember that this generated page is just a starting point. It’s fully editable. You can change colors, resize visuals, or tweak the charts just as you would any other Power BI report.

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

Iterating and Refining Your Page

The initial output might be good, but it's rarely perfect. The real benefit of Copilot is the ability to have a follow-up conversation to refine the report.

Looking at the page it created, you might have new questions. You can give follow-up commands in the Copilot pane:

  • "Change the bar chart to show cost per conversion by channel."
  • "Add a slicer for campaign date."
  • "Show me the top 5 campaigns by conversion rate in a table."
  • "Add a map visual showing conversions by geographic location."

With each prompt, Copilot modifies the report page, adding, removing, or changing visuals. This conversational approach turns hours of manual adjustments into a few seconds of typing.

Going Deeper: Writing DAX and Summarizing Data

Beyond building visuals, Copilot is an incredibly powerful tool for two other data analysis tasks: writing complex DAX measures and generating narrative summaries of your insights.

Creating DAX Calculations with Copilot

DAX can feel like learning a new programming language. Copilot can handle the heavy lifting for you. You don't create DAX measures directly in the main chat pane, instead, you can suggest adding them as part of your page creation prompts, or you can go to the calculation view to get suggestions.

Let's say your dataset has 'Total Sales' and 'Total Costs' but not 'Profit Margin'. You could simply ask:

"Create a DAX measure for Profit Margin."

Copilot generates the formula, which you can then add to your data model and use in your reports.

Profit Margin = DIVIDE([Total Profit], [Total Revenue])

This capability is a huge accelerator for both beginners who are unfamiliar with DAX conventions and for experts who want to automate the creation of standard calculations.

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.

Generating Data Summaries

Once your report page is complete, Copilot can help you interpret it. In the Visualizations pane, there’s a new "Narrative with Copilot" visual type. Adding this to your report will generate a text-based summary of the insights present on the page.

It might produce bullet points like:

  • Overall revenue increased by 12% in the last quarter, primarily driven by sales in the "Electronics" category.
  • The campaign "Summer Sale 2024" had the highest return on ad spend at 5.2x.
  • While North America remains the largest market, there's been a significant 30% month-over-month growth in the European market.

This feature is excellent for quickly pulling out the "so what?" from your data. You can copy this summary into an email for stakeholders or use it as a foundation for your own deep-dive analysis. Again, it’s a starting point - you should always validate the insights and add your own human context.

Best Practices for Effective Copilot Prompts

The quality of your output depends heavily on the quality of your input. Like any AI tool, vague prompts lead to vague results. Here’s how to communicate with Copilot effectively:

  • Be Specific and Clear: Instead of "Show me sales data," try "Create a line chart showing our total revenue by month for the last fiscal year."
  • Provide Context: Include details about timeframes, categories, or specific metrics you want to see. For example, "...for our top 10 products" or "...filtered only to the 'US' region."
  • Use the Names from Your Data Model: If you know your column names (e.g., 'SalesAmount', 'CustomerSegment'), use them in your prompt. This helps Copilot find the right data without guessing.
  • Start Broad, Then Narrow Down: It’s often easier to ask for a general report page first and then use follow-up prompts to refine it, rather than trying to build the perfect page with one massive, complex prompt.
  • Review and Verify: Always double-check what Copilot creates. It's smart, but it can still misunderstand nuances in your data model or intent. Trust but verify.

Final Thoughts

Power BI Copilot is a significant step toward making business intelligence more intuitive and conversational. By handling the tedious technical work of report building and DAX writing, it allows you to focus less on the 'how' of data analysis and more on the 'what' and 'why' - uncovering the actual insights that drive better decisions.

The philosophy behind tools like Power BI Copilot is exactly what most marketing and sales teams need: less complexity, faster insights. At Graphed , we’ve applied this same natural language approach to the platforms you use every day, like Google Analytics, Shopify, Facebook Ads, and HubSpot. While Power BI is a powerful, general-purpose tool, our platform is built specifically to connect your marketing and sales data sources with a few clicks, letting you build real-time, cross-platform dashboards just by asking questions - no setup, data modeling, or admin configurations required.

Related Articles