As organizations become increasingly data-driven, the ability to quickly interpret and communicate insights from raw information is becoming an essential skill. The Analyst agent in Microsoft 365 Copilot enables users to do just that by analyzing and visualizing data directly within familiar tools like Excel and Forms. This lab provides a hands-on demonstration of how to use the Analyst agent to understand existing datasets—whether from surveys, spreadsheets, or poll results—and transform them into actionable insights with minimal effort.

In this scenario-based exercise, you’ll learn how to use the Analyst agent to explore trends, identify anomalies, and generate visuals that enhance data storytelling. Whether you’re reporting on team performance, customer feedback, or operational metrics, this lab shows you how to turn raw numbers into clear summaries or charts ready to share with stakeholders. It’s a powerful way to save time, reduce manual effort, and build confidence in analytical decision-making with AI support.

Exercise

You’ve been tasked with analyzing the results of a survey related to an internal initiative called “Project Nexus,” a six-week pilot program aimed at improving cross-department collaboration through a new digital workspace platform. The project involved employees from various departments—including IT, HR, Marketing, and Operations—who were invited to use the platform for daily communication, document sharing, and task management. The goal of the project was to assess the platform’s effectiveness in terms of productivity, communication flow, and meeting deadlines.

At the end of the pilot, participants were surveyed to evaluate their satisfaction with the project, the clarity and effectiveness of communication, adherence to the proposed timeline, and their overall experience with the new system. You plan to use the preconfigured Analyst agent in Microsoft 365 Copilot to explore the Project Nexus survey results. As with any preconfigured agent, you can enter your own custom queries or use the starter queries suggested by the agent. These queries are designed to produce quantitative and qualitative analyses, visualizations, and insights and recommendations about the project.

Steps to guide the Analyst agent in interpreting and visualizing the Project Nexus survey results:

1. Select the following link to download a copy of the Project Nexus survey results. Click the Download button at the top of the screen to save the file to your device.
2. In Microsoft Edge, open a new tab and enter the following URL: https://M365copilot.com
3. In Microsoft 365, select the Analyst agent if it appears in the navigation pane under the Agents section. If not, select All Agents in the navigation pane, then in the Agent Store window, select Analyst under the Created by Microsoft section.
4. In Microsoft 365, the Analyst agent window will appear. In the input field, click the Add content and agents icon, represented by a plus sign (+).
5. In the menu that appears, select Upload from this device. In File Explorer, go to the Downloads folder, select the Project Nexus Survey Results file you downloaded, and click Open.
6. In the input field, enter the following query next to the linked file:
   Analyze this table and tell me the top three trends.

You want to dive deeper into each category, so start by entering the following query:

What is the average rating for each survey category?

During our tests, the agent returned what appeared to be a blank page. In reality, it’s not empty—there’s simply a large blank space between the agent’s response and the input field. If this happens, scroll up using the vertical scrollbar to find the response. And if you scroll all the way down, the input field should reappear. This agent is new, so this blank space seems to be a bug. That said, by the time you complete this exercise, the issue may be resolved. In any case, review the results. If the Analyst agent suggests a next step, such as “Would you like a visual comparison of these averages?”, enter Yes in the input field (if the blank space issue persists, you’ll need to scroll to the bottom to see the input field).

Again, scroll up to review the previous query results (if needed) and examine them.
At this point, you can spend as much time as you like analyzing the survey results with the Analyst agent.

You can enter your own custom queries or try one of the following depending on the type of analysis you want:

Quantitative Analysis Queries:

• Which category received the highest average rating, and which the lowest?
• How many participants rated project satisfaction 4 or higher?
• What percentage of participants rated timeline adherence below 3?
• Can you identify correlations between communication effectiveness and overall experience?

Qualitative Analysis Queries:

• Summarize the most frequent themes in the comments section.
• Are there recurring concerns or suggestions mentioned in the comments?
• Identify comments that mention communication or timeline issues.

Insights and Recommendations Queries:

• Based on the survey data, what are the top three strengths of Project Nexus?
• What are the main areas for improvement suggested by participants?
• Provide a summary report of the survey results with actionable recommendations.

Quantitative Visualization Queries:

• Generate a pie chart of overall rating distribution.
• Create a bar chart comparing average ratings for project satisfaction, communication effectiveness, timeline adherence, and overall experience.
• Plot a histogram of satisfaction ratings to visualize their distribution.
• Generate a scatter plot to analyze the relationship between communication effectiveness and overall experience.
• Create a heatmap of correlations across all numeric rating categories.
• Create a box plot for each rating category to show range and quartiles.
• Plot a line chart showing timeline adherence ratings per participant, sorted by ID.

Next Unit: Review the Researcher Agent