Summary  
This chapter covers generating and applying substring extraction logic via an AI coding assistant to split text fields on a delimiter and populate a new column, illustrating the difference between simple pattern matching and using live data access.  

General domain of usage  
Spreadsheet data processing



In this chapter, you start working with real data and use Copilot to extract useful information.

The goal is to fill the Sales Channel column based on the Order ID. Each Order ID contains a prefix that indicates where the sale happened, so you need to extract the text before the first hyphen.



Instead of searching for formulas manually, you can ask Copilot directly.

Describe the task clearly and ask for the formula as output. This keeps you in control and helps you understand what is being applied.

Use Copilot as a Formula Assistant


When you use Copilot in chat mode, it behaves like a general assistant. It does not reliably see your actual data.

Because of this, it may guess column names or return formulas that do not match your table.

Copilot Without Edit Mode


When you enable Edit with Copilot, it can access your workbook directly.

It reads your tables, detects column names, and generates formulas that match your data. This makes the results more accurate and reduces manual corrections.

Copilot with Edit Mode


Once you have the correct formula, insert it into the Sales Channel column.

Because you are working with a table, Excel fills the formula automatically for all rows.

Apply the Formula

Discover how to use Excel with Copilot to solve real business problems. Explore practical workflows that turn messy data into structured, reliable systems. Learn how to build models, automate calculations, and generate insights without memorizing formulas. Understand how to guide Copilot, validate its output, and keep full control over your logic.

Explore the core setup and habits for working with Copilot in Excel. You configure your environment, learn how to write clear prompts, and apply safe workflows before using Copilot on real data. You focus on testing, verifying results, and protecting your files. These skills help you avoid mistakes and use Copilot with confidence further.

Focus on turning messy, inconsistent datasets into clean and usable data. You standardize headers, split combined fields, fix mixed formats, and remove duplicates using Copilot. Consider real-world scenarios and apply structured workflows to clean data safely.

Discover using Copilot to work with clean data and turn it into insights. You extract values, connect multiple datasets, handle errors, and build dynamic summaries. Learn to describe tasks in plain language while Copilot generates and applies the formulas.

Learn how to present data in a clear and interactive way. Create PivotTables, add slicers, build charts, and apply conditional formatting to highlight key insights. Combine all elements into a structured dashboard and use Copilot to support analysis, formatting, and presentation. 

Build a forecasting model using assumptions, pricing scenarios, and 12-month projections. Test different outcomes, guide Copilot through complex logic, and turn your analysis into a clear, decision-ready recommendation.

Turn your work into a reusable, product-like spreadsheet. Clean raw data, combine multiple sources into a single system, build automated reports, and package everything into a polished template with instructions that anyone can use.

How Copilot Sees Your Data