Use Excel Copilot to Analyze Risk Data

What This Does

Copilot in Excel helps you analyze risk data using natural language. asking for pivot analyses, trend charts, and formulas without needing to build them manually. Useful for summarizing loss event data, KRI trend analysis, and risk scoring.

Before You Start

• You have Excel open (desktop or web at excel.microsoft.com) with your risk data in it
• You're signed in with a Microsoft 365 Business or Enterprise account with Copilot enabled
• Your columns have clear headers (Date, Risk Category, Loss Amount, Business Unit, Control Status)

Steps

1. Open Copilot

Click the Copilot icon in the lower-right corner of the Excel window. The panel opens on the right side of your spreadsheet.

Microsoft consolidated every way of opening Copilot into this one corner button, so it is no longer on the Home ribbon. Keyboard shortcuts: Alt + C on Windows, Cmd + Ctrl + I on Mac, or F6 anywhere. If you prefer the old ribbon placement, right-click the corner button and choose Move to ribbon.

2. Give your columns clean headers

Copilot reads your headers to figure out what it is analyzing, so "Loss_Amount" works and "Col_D" does not. Formatting the range as a table (select it, press Ctrl+T) is still worth doing because it keeps ranges stable as rows are added. It is no longer a hard requirement, though: Copilot now reads ordinary ranges too.

3. Ask questions in plain English

Type your analysis request:

• "Summarize total losses by risk category for Q1 2026 and show the top 5 business units by loss amount"
• "Show a trend chart of KRI breach count by month over the last 12 months"
• "Flag any rows where loss amount exceeds $100,000 and control status is 'ineffective'"
• "Calculate the percentage change in operational loss events from Q4 2025 to Q1 2026 by category"

Copilot generates the analysis, chart, or formula and inserts it into your spreadsheet.

4. Review the output

Check the generated pivot tables and charts against your raw data for accuracy. Copilot can occasionally misinterpret column names, if results look wrong, rephrase the question with more specific column names.

Real Example

Scenario: You have a loss event database with 200 rows from the past year. You need to prepare a summary for the Operational Risk Committee showing total losses by category and quarter.

What you type: "Create a pivot analysis showing total loss amount by risk category (columns) and quarter (rows). Also add a column for number of events."

What you get: A pivot table is inserted into a new sheet with exactly that structure, ready to paste into your report.

Tips

• Always verify Copilot's calculations by spot-checking a few rows manually, especially for financial data going to senior management or regulators
• If Copilot doesn't understand your column names, rename them to something clearer first (e.g., "Loss_Amount" instead of "Col_D")
• Use "Add a column" requests to generate formula columns: "Add a column that calculates the residual risk score as inherent score minus control effectiveness score"
• The pane has a model picker in its upper-right corner. Auto is fine for summaries and charts. For a gnarly multi-step calculation, pick the reasoning model and check its work.

Tool interfaces change, if a button has moved, look for similar AI/magic/smart options in the same menu area.