Excel Copilot prompts: 25 practical examples for everyday office work

Use 25 practical Excel Copilot prompts for formulas, cleanup, analysis, charts and reports. Learn a repeatable framework and checks for accurate results.

Good excel copilot prompts tell Copilot four things: what data to use, what task to perform, what rules to follow, and what output you want. Instead of “analyze sales,” ask: “Using the Sales table, summarize Revenue_AED by Region for the last three months, flag any month more than 10% below the previous month, and return a PivotTable plus three bullet insights. Do not change the source table.” Microsoft’s current guidance is similar: be specific, name columns or ranges, refine with follow-up questions, and verify the result. The 25 prompts below are designed for routine office work, not demos.

How to write a strong Excel Copilot prompt

Use a four-part pattern: scope + task + rules + output. Scope identifies the table, range or columns. Task states the job: calculate, clean, compare, summarize, chart or explain. Rules define conditions and exclusions. Output specifies a formula, PivotTable, chart, summary or answer only.

For example, “Find late invoices” is vague. A stronger version is: “In the Invoices table, flag rows where Status is not Paid and Due_Date is earlier than today. Add a new column called Follow_Up with Overdue or Current. Do not change Invoice_ID or Amount_AED.” The second prompt gives Copilot a data source, condition, output and protection rule.

Before prompting: prepare the workbook

Copilot works best with structured data: clear headers, consistent data types and one logical record per row. Microsoft’s direct-analysis documentation supports tables and table-like ranges, including multiple ranges, but not unstructured data. Avoid merged headers, unexplained subtotals and mixed-purpose columns.

Also decide whether you want Copilot to analyze or edit. Microsoft documents a chat-style experience that can return insights without modifying the workbook, while its editing experience can make workbook changes using Excel features such as tables, formulas, charts, PivotTables, formatting and validation. If preservation matters, say “answer only; do not modify the workbook.” If you want action, say exactly what should be added or changed.

25 practical Excel Copilot prompts

Data cleanup and quality checks

1. Duplicate check — “In the Contacts table, identify duplicate rows using Email as the key. Create a review list showing the duplicate Email, Name and Company. Do not delete any rows.” This is safer than asking Copilot to remove duplicates immediately because a human can review the matches first.
2. Missing-data audit — “Check Orders for blanks in Order_Date, Customer, Region, Revenue_AED and Salesperson. Create a summary showing the number and percentage of missing values in each column, then highlight affected rows. Do not fill missing values automatically.”
3. Standardize categories — “Review the Region column and identify inconsistent labels that appear to mean Dubai, Abu Dhabi or Sharjah. Propose a standardized mapping in a separate table before changing the source values.”
4. Number-as-text check — “Find values in Amount_AED that are stored as text or contain non-numeric characters. List the affected row IDs and suggest the safest conversion. Do not alter the original column.”
5. Date-quality check — “Review Invoice_Date and Due_Date. Flag invalid dates, Due_Date earlier than Invoice_Date, and dates outside 2026. Add a Reason column explaining each flag.”

Formulas and calculations

6. Conditional formula — “Add a column called Budget_Status. Return Over Budget when Actual_AED is greater than Budget_AED, Within 5% when Actual_AED is within 5% of Budget_AED, and On Track otherwise. Explain the formula before inserting it.”
7. Lookup — “In the PurchaseOrders table, add Current_Stock by looking up SKU against the Inventory table. Use the most appropriate Excel lookup formula, return Not Found when no SKU matches, and explain why you chose that formula.”
8. Days outstanding — “Add Days_Outstanding using Invoice_Date and Payment_Date. If Payment_Date is blank, calculate through today. Return blank when Invoice_Date is missing. Format the result as a whole number.”
9. Percentage contribution — “Add a column showing each department’s Expense_AED as a percentage of total expenses. Format as a percentage with one decimal place and add a total check that equals 100%.”
10. Explain inherited logic — “Explain the formula in cell H27 in plain English. List every referenced range, describe what happens when a lookup fails, and tell me one edge case I should test. Do not change the formula.” Microsoft specifically supports asking Copilot to explain existing formulas.

Summaries, PivotTables and analysis

11. Monthly expense summary — “Using Expenses, create a PivotTable on a new sheet showing total Amount_AED by Department in rows and Month in columns. Sort departments by annual total descending and include grand totals.”
12. Sales performance — “Compare Revenue_AED by Region for the last six complete months. Show month-over-month percentage change and identify the two largest declines. Answer with a table and three concise observations; do not modify the source.”
13. Top-customer concentration — “Calculate what percentage of total Revenue_AED comes from the top 5 customers. Show the five customers, their revenue, cumulative share, and the remaining-customer share.”
14. Budget variance — “For each Department, compare Actual_AED with Budget_AED. Show variance in AED and percentage, rank the five largest unfavorable variances, and create a short management summary.”
15. Outlier review — “Identify unusually high Expense_AED transactions by Department. Show the method you used to define an outlier, list the affected rows, and do not remove them.” This makes the analytical assumption visible instead of hiding it behind a result.

Charts and visual reporting

16. Trend chart — “Create a line chart of monthly Revenue_AED for 2026 with one series per Region. Add a clear title and keep the source table unchanged. After creating it, summarize the two most important trends.” Microsoft currently supports requesting charts and specifying the desired visual.
17. Budget comparison chart — “Create a clustered column chart comparing Budget_AED and Actual_AED by Department. Sort departments by Actual_AED descending and place the chart on a new sheet called Budget Review.”
18. Conditional formatting — “Highlight rows where Days_Outstanding is over 45 and Balance_AED is above 10,000. Use one consistent conditional-formatting rule and explain the condition before applying it.”
19. Executive KPI view — “Create a compact summary showing Total Revenue_AED, Total Cost_AED, Gross Profit_AED and average order value, then add one appropriate chart showing monthly revenue. Keep the summary on a new sheet called Management Summary.”
20. Inventory risk visual — “Create a chart showing SKUs where Stock_On_Hand is below Reorder_Level, grouped by Category. Include only items that need attention and add a table with SKU, stock, reorder level and supplier.”

Text, reporting and office administration

21. Feedback themes — “Analyze the Customer_Comment column. Group comments into no more than six practical themes, count comments per theme, summarize sentiment, and add a Theme column. Keep the original comment unchanged.” Microsoft documents Copilot analysis of text columns for themes and sentiment.
22. Management commentary — “Using the monthly KPI table, write five bullet points for a management review. Each bullet must cite a specific figure from the workbook, distinguish positive movement from risk, and avoid guessing causes that are not supported by the data.”
23. Action list from tracker — “Review the ProjectTracker table. Create an action list for items where Status is not Complete and Due_Date is within the next 7 days or already overdue. Show Owner, Task, Due_Date and Priority, sorted by urgency.”
24. Vendor review — “Summarize vendor performance using Supplier, Spend_AED, Delivery_Days and Defect_Rate. Rank suppliers using transparent criteria, show the criteria and weights first, and do not create a final score until I approve the weighting.”
25. Workbook audit — “Review this workbook for practical reporting issues: inconsistent headers, obvious blank-data problems, formulas with errors, sheets with unclear names, and charts without descriptive titles. Return a prioritized checklist only. Do not make changes.”

Prompt checklist: what to include

How to improve a weak prompt step by step

Start with the business question, not the Excel feature. Instead of “make a PivotTable,” write “I need to see which departments caused the increase in monthly expenses.” Then add the data scope: “Use Expenses, especially Department, Date and Amount_AED.” Add a rule: “Exclude Status = Cancelled.” Finally specify the output: “Create a PivotTable by month and department and give me three observations.”

Use follow-up prompts rather than packing every instruction into one message. Microsoft recommends refining results. For example: “Now show only departments with more than AED 25,000 variance,” then “Create a chart for those departments,” and finally “Explain the largest change using only workbook evidence.”

Original office example: monthly procurement review

Assume a UAE office has a PurchaseOrders table with 1,800 rows and columns PO_Date, Supplier, Category, Quantity, Unit_Cost_AED, Total_AED, Delivery_Days and Status. The procurement manager wants to prepare a monthly review without manually rebuilding filters and summaries.

A useful first prompt is: “Using PurchaseOrders, summarize Total_AED by Supplier and Category for August 2026. Exclude Status = Cancelled. Show the top 10 suppliers by spend, average Delivery_Days for each, and flag any supplier whose average delivery time exceeds 10 days. Return a PivotTable and a short summary; do not change the source table.” Suppose Copilot reports AED 684,000 total eligible spend and identifies three high-spend suppliers with average delivery above 10 days.

Do not copy those figures straight into management reporting. Reconcile the AED 684,000 against a standard PivotTable or SUMIFS-based check, inspect the excluded statuses, and spot-check supplier averages. Then follow up: “For the three flagged suppliers, show the five largest POs and whether late delivery is concentrated in one Category.” The sequence turns a generic AI request into an auditable office workflow.

Common mistakes with Excel Copilot prompts

The biggest mistake is being vague: “clean this” or “analyze everything.” Another is omitting column names when similar fields exist. Also avoid asking Copilot to delete or overwrite data before reviewing what it found; ask it to flag or propose changes first.

Do not assume every Copilot response changes the workbook. Microsoft’s direct-analysis experience can answer questions and return static tables or images without modifying source data, while editing features are designed to act on the workbook. State your intent. Also avoid building a new long-term workflow around the experimental =COPILOT worksheet function: Microsoft says it will no longer be available starting September 14, 2026.

Finally, verify generated work. Microsoft repeatedly advises users to review AI-generated content. For financial or operational reporting, check totals, filters, dates, formulas and assumptions. Copilot can speed up the first draft of analysis; accountability for the final workbook remains with the person using it.

Frequently asked questions

What are the best prompts for Copilot in Excel?

The best prompts are specific to the workbook. Name the table or columns, state the task, define rules and request a clear output. “Using Sales, compare Revenue_AED by Region for Q2 and create a PivotTable” is stronger than “analyze sales.”

How do I get better results from Excel Copilot?

Use clear headers and structured data, name the columns in the prompt, state whether Copilot may edit the workbook, and refine the first result with follow-up questions. Microsoft’s current tips explicitly recommend specificity and naming relevant columns or ranges.

Can Copilot create formulas in Excel?

Yes. Microsoft documents Copilot generating formula columns, summary rows and single-cell formulas, as well as explaining existing formulas. Review the suggested formula and test important cases before relying on it.

Can Copilot create PivotTables and charts?

Yes. Current Microsoft documentation shows Copilot creating charts and PivotTables from workbook data. You can specify the measure, category, time period and desired output in the prompt for a more useful result.

Can Copilot analyze text in Excel?

Yes. Microsoft documents analysis of text such as survey responses, reviews and feedback, including summarizing comments, identifying themes and sentiment, and adding results to the workbook. Keep the original text so classifications can be reviewed.

Does Copilot in Excel change my workbook automatically?

It depends on the experience you use. Chat/direct-analysis scenarios can provide answers without modifying the workbook, while editing experiences can make changes to sheets, ranges, formulas, formatting and other workbook elements. State “do not modify the workbook” when you only want analysis.

Is the =COPILOT function the same as Copilot in Excel?

No. The =COPILOT worksheet function is a separate experimental formula experience. Microsoft currently states that it will no longer be available starting September 14, 2026. The main Copilot in Excel experience remains available subject to eligible licensing and organizational settings.

Conclusion

Effective excel copilot prompts are clear work instructions. Specify the data, task, rules and expected output. Start with reviewable tasks such as data-quality checks, formulas, summaries, PivotTables and charts, then refine with follow-up prompts. Verify important totals and assumptions before sharing. If the source file is messy, use the relevant Excel or CSV cleanup utility on ULoveOffice.com first, then bring the cleaner workbook into Copilot and use one of the 25 prompts above.

Internal-link suggestions

• How to use Python in Excel with Copilot for advanced analysis — deeper follow-on for users who outgrow everyday prompts.
• 10 Excel formulas every office manager should know in 2026 — useful when a normal formula is simpler than Copilot.
• CSV opens wrong in Excel: encoding fixes for Arabic text — relevant before prompting against imported Arabic/UAE data.
• Timesheet template: daily, weekly and monthly formats — a practical office dataset for formula, summary and action-list prompts.
• Markup vs margin: the difference with worked examples — useful for checking profitability prompts and formula outputs.

ULoveOffice.com CTA recommendation

Primary CTA: the most relevant live Excel data-cleaning, CSV-to-Excel, Excel repair, or spreadsheet-preparation utility on ULoveOffice.com. Suggested anchor: “Clean your Excel data before using Copilot.” Insert the exact live URL only after verifying the current tool.

Image brief

Hero — 1200×630
Filename: excel-copilot-prompts-25-examples-hero.jpg
Brief: Editorial illustration of an office spreadsheet with a generic AI assistant pane and several prompt cards for formulas, cleanup, PivotTables, charts and reporting. Show AED values and a professional office-data feel. Avoid copying Microsoft UI exactly and avoid trademark logos.
Alt text: Excel Copilot prompts for formulas, analysis, charts and office reporting

Supporting screenshot 1
Filename: excel-copilot-prompt-anatomy.png
Brief: Annotated live Excel screenshot showing a prompt broken into Scope, Task, Rules and Output. Use a simple Expenses table with AED values.
Alt text: Four-part structure of an effective Excel Copilot prompt

Supporting screenshot 2
Filename: excel-copilot-pivottable-prompt.png
Brief: Live Copilot pane with the procurement-review prompt and resulting PivotTable or proposed action.
Alt text: Excel Copilot prompt creating a procurement PivotTable

Supporting screenshot 3
Filename: excel-copilot-formula-explanation.png
Brief: Select an existing formula cell and show Copilot explaining the formula in plain English.
Alt text: Copilot in Excel explaining an existing formula

Supporting screenshot 4
Filename: excel-copilot-verify-results.png
Brief: Copilot-generated summary beside a simple Excel reconciliation check, emphasizing verify-before-share workflow.
Alt text: Checking Excel Copilot results against workbook totals

Sources

• Microsoft Support — Copilot in Excel tips (effective prompts, specificity, column/range naming and follow-up refinement), accessed 5 September 2026.
• Microsoft Support — Get started with Copilot in Excel (editing and chat capabilities, licensing/availability), accessed 5 September 2026.
• Microsoft Support — Get data insights with Copilot in Excel (insights, formulas, lookups, text analysis and formula explanations), accessed 5 September 2026.
• Microsoft Support — Visualize your data with Copilot in Excel (charts, PivotTables, highlighting, sorting and filtering), accessed 5 September 2026.
• Microsoft Support — Get direct answers to your data analysis questions (tables/table-like ranges, analysis scope and non-editing behavior), accessed 5 September 2026.
• Microsoft Support — Use data from the web and internal sources with Copilot in Excel, accessed 5 September 2026.
• Microsoft Support — Use skills with Copilot in Excel to complete repeatable tasks, accessed 5 September 2026.
• Microsoft Support — COPILOT Function (notice that the experimental worksheet function will no longer be available starting 14 September 2026), accessed 5 September 2026.

Editorial verification date: 5 September 2026. Microsoft Copilot features, licensing, entry points and UI labels can change. Re-check Microsoft documentation before publication. This article makes no UAE legal, tax or HR claims; AED values are original office-work examples only. The article was also checked against recovered ULoveOffice topics including Python in Excel with Copilot, Excel formulas, CSV Arabic encoding, timesheets, markup vs margin and working-days content to avoid substantial duplication.