chatgpt prompts for excel: 7 practical templates to use
Use chatgpt prompts for excel to draft formulas, plan PivotTables and review weekly reports. Copy seven practical templates and check each result before use.

chatgpt prompts for excel: 7 practical templates to use
Good chatgpt prompts for excel specify the columns, calculation rules and output you need. Choose one template below, replace the bracketed details and check the result against a small example before using it in your working spreadsheet.

Start with a task you already understand. If you prepare the same sales summary every Friday, use that report for your first trial. You should be able to explain what a correct answer looks like before asking for help.
Which spreadsheet task should you give AI first?
Choose a formula, a summary or a report outline with a clear finish line. For the ChatGPT templates here, request written instructions and a proposed calculation. For work inside Excel, specify exactly which worksheet Copilot should change.
Copilot in Excel is an AI assistant that helps build and edit workbooks from written requests. Its capabilities include formulas, charts and PivotTables. A PivotTable is a summary that groups records into categories such as products or weeks. See Microsoft's explanation.
| Your task | Ask for | Check yourself |
|---|---|---|
| Calculate order totals | A formula with assumptions | One complete order |
| Compare sales channels | A grouped summary | Included statuses and totals |
| Prepare a team update | A short report from verified figures | Every number and explanation |
Keep interpretation separate from calculation. First establish what happened in the spreadsheet. Then decide which questions deserve investigation. If your next step involves customer conversations, use the separate sales prompt templates after checking the figures.
What should your prompt include before you paste any data?
Describe what one row represents, name the columns and define the exceptions. A prompt is simply your written instruction to the assistant. Treat it like a brief for someone who has never seen your spreadsheet.
Microsoft recommends naming relevant columns or ranges and refining requests with follow-up details. See its Copilot prompting tips.
- Workbook context: an order export, project tracker or campaign report.
- Row meaning: one order, one order item or one payment.
- Column names: copy the actual headings.
- Business rules: which statuses and dates count.
- Output: a formula, instructions or a proposed summary.
- Acceptance check: a sample with a known expected result.
Use invented customer details for your first trial. Remove information the task does not need, and work from a copy of your file. Write down how you want blanks, cancellations and refunds handled before sending the prompt.
Which seven prompts can you copy and adapt?
The examples below are hypothetical instructions, not reports of tested business results. Replace every bracketed item. Start with one template and keep the sample small enough to inspect manually.
1. Calculate the amount collected
In my Orders worksheet, B is Quantity, C is Unit price and D is Discount amount for the entire row. Data starts in row 2. Propose a formula for E2 that multiplies quantity by price and subtracts the discount. Treat a blank discount as zero. Leave the result blank if quantity or price is missing. Explain the formula and provide invented test rows with expected results.
Before using the answer, confirm whether your discount really is a money amount. If it is a percentage, change the brief. Check one ordinary row, one discounted row and one incomplete row.
2. Repair a formula without rebuilding the sheet
My Excel version is [version]. This formula produces [exact error or result]: [formula]. The relevant columns are [headings]. With these invented inputs [values], I expect [result]. Ask for missing details before proposing a fix. Give the smallest correction, explain what changed and show how I can test it. Keep my existing column layout.
Copy the actual formula and error message. Avoid replacing an unknown detail with a guess. Test the correction in one cell before extending it to the rest of the worksheet.
3. Review inconsistent category names
Review this invented list of expense categories: [list]. Suggest which labels might refer to the same category. Return original label, suggested label and a question for the reviewer. Keep subscriptions, equipment purchases and delivery charges separate unless I approve a different grouping. Flag uncertain matches. Produce a review table before suggesting any changes to the source.
Read each proposed grouping before accepting it. Keep a short list of approved category names for the next import. For the business decisions after that cleanup, continue with the guide to reviewing small business costs.
4. Plan a sales PivotTable
My worksheet has Order date, Channel, Order ID, Status and Collected amount. Each row represents one order. Plan a PivotTable for [start date] through [end date], including only Completed orders. Put Channel in rows and the sum of Collected amount in values. List incomplete records separately. Explain how to reconcile the total with the source rows using the same filters.
Reconcile means checking that two totals agree. Choose one channel and inspect its source records manually. If each row actually represents an order item, rewrite the row definition before requesting order counts.
5. Design a useful campaign chart
I want to compare weekly enquiries across marketing channels. My columns are Week starting, Channel and Enquiries. Suggest a chart and describe the summary table it needs. Label incomplete weeks clearly. Do not fill missing observations with invented values. Include the period and metric in the title. Explain what this chart cannot tell us about sales or profit.
Keep the question narrow: enquiries by channel. Copilot can create charts and PivotTables within Excel; see Microsoft's capability guide. For a wider channel review, use the social media audit guide.
6. Write a report from checked figures
Use only this verified summary table: [table]. Write a short weekly update for a small agency owner. Include the reporting period, the main changes and questions needing investigation. Attach the supporting value to every observation. Separate observations from suggested actions. Do not invent reasons for a change. If a comparison period is missing, say that comparison is unavailable.
Check the numbers before editing the tone. If the report raises a content question, use Instagram prompt templates for that next task. Keep any proposed explanation labelled as a hypothesis until you investigate it.
7. Plan a repeatable reporting process
Each week I receive an order export with [headings]. Propose a repeatable Excel reporting checklist covering import review, status filtering, channel totals and a short written update. For each stage, list the input, expected output and reviewer check. Define what should happen if columns change, dates are missing or totals disagree. Give me the plan first, with no changes to the workbook.
Use this as a process brief. Do not treat a written plan as an installed automation. Before considering repeated execution, run the same checklist on another sample file and record what needed correction.
How do you check an answer before relying on it?
Inspect source rows and calculations before accepting a polished summary. Microsoft warns that Copilot output can be inaccurate and should be reviewed and verified. See its accuracy guidance.
- Calculate one ordinary example independently.
- Check a blank input and confirm the intended response.
- Inspect a cancelled or refunded record.
- Compare totals using identical dates and filters.
- Trace each written observation back to a supplied value.
If something fails, name the mismatch precisely. For example: “The summary includes cancelled orders; use Completed only and recalculate.” Keep the original sample so you can tell whether the correction actually resolved the problem.
For a cash-related report, label collected payments separately from order values. Decide which measure your report needs before drafting it. Use the cash flow planning guide when moving from spreadsheet preparation to a wider business review.
How can you turn a good prompt into a weekly routine?
Save the prompt alongside its assumptions and checking steps. Give the report a clear owner. Record where the input comes from and what should stop the process.
Consider a hypothetical freelance studio preparing a Friday project summary. Ask the person running it to record preparation, correction and review time separately. On the next run, use the same report definition and compare the work required.
Keep a brief handover note: accepted headings, date boundaries, excluded statuses and unresolved questions. Have a colleague follow it once. Improve any step they cannot complete without asking you what you meant.
What do business owners ask about Copilot in Excel?
What is Copilot in Excel?
It is an AI assistant for building and editing Excel workbooks. Start with one clearly defined worksheet task. Microsoft FAQ.
What can Copilot in Excel do?
It can work with formulas, charts, PivotTables and formatting. Choose one output for your first trial. Capabilities.
How do I start using Copilot in Excel?
Select the Copilot icon in Excel's lower-right corner. If it is missing, check your subscription and organization settings. Getting started.
How do I write a good prompt for Copilot in Excel?
Name the columns or range, define the calculation and specify the output. Add missing detail in a follow-up request. Prompting tips.
Which sources should you check for product details?
- Microsoft: Get started with Copilot in Excel. Check capabilities and opening instructions.
- Microsoft: Copilot in Excel tips. Check guidance for refining requests.
- Microsoft: Frequently asked questions. Check access requirements and limitations.
Follow me on Instagram
Short notes, practical examples and daily digital strategy ideas.
I'm Anar Rustamli - a strategist, entrepreneur, and AI adoption leader working at the edge of growth, technology, and human thinking. Since 2016, my work has focused on helping businesses evolve in a rapidly changing digital landscape. I design growth systems, AI-powered workflows, and strategic frameworks that align performance with purpose. I believe real growth happens when strategy, data, and human insight work together - and my mission is to help businesses adopt AI in a way that strengthens both their results and their identity.

