Preparing your data (“Scrubbing”) for use with Copilot

Owned by Laurie Fox, created
Jan 07, 2025
2 min read

This article will clarify the concept of data scrubbing and provide a step-by-step guide on how to effectively perform it.

To scrub data before using it with Copilot, you need to clean and format your data by removing duplicates, handling missing values, ensuring consistent formatting, and removing sensitive information; you can utilize features like Excel's "Clean Data" function powered by Copilot to automate some of this process, depending on your data source and the specific issues you need to address. 

 

Why is data scrubbing important for Copilot?

  • Accurate results:

    Clean data ensures Copilot generates accurate and reliable insights, as it is trained on high-quality information. 

  • Improved performance:

    By removing inconsistencies and errors, you minimize the workload for Copilot, allowing it to focus on generating relevant and meaningful outputs. 

  • Ethical considerations:

    Scrubbing sensitive data is crucial to protect privacy and comply with data protection regulations. 

 

Key steps for scrubbing data for Copilot:

  • Identify data quality issues:

    Review your data to identify inconsistencies, missing values, incorrect data types, duplicates, outliers, and any sensitive information that needs to be removed. 

  • Remove unnecessary columns:

    Eliminate columns that are not relevant to the analysis or could introduce noise in the Copilot results. 

  • Handle missing data:

    Decide how to handle missing values, such as filling them with a placeholder value, removing rows with missing data, or imputing values based on other data points. 

  • Normalize data:

    Ensure consistent formatting across your data, including consistent capitalization, date formats, and decimal points. 

  • Remove duplicates:

    Identify and remove duplicate rows or entries to avoid redundancy in your data. 

  • Anonymize sensitive data:

    If necessary, remove personally identifiable information (PII) or other sensitive details before feeding data to Copilot. 

  • Use data cleaning tools:

    Leverage built-in data cleaning features in your spreadsheet application like Excel's "Clean Data" function powered by Copilot, which can automatically detect and fix common data issues with a single click. 

  

More information: How to Prepare Data for Microsoft 365 Copilot - Shelf

Related articles