Attribution: Prepared by 6DuckLearn. Inspired by Google Workspace with Gemini Prompting Guide 101 (2025).
When to use this
You've exported your employee survey responses into a spreadsheet and the data is messy — blank names, inconsistent region labels, and rows with no satisfaction score. You need a clean file before you can analyse or report on results.
What you'll get
A cleaned copy of your survey spreadsheet with: blank names replaced by "Anonymous", normalised region/location labels, blank satisfaction rows removed, and a new file generated — leaving the original untouched.
Prompt template
Use this in the Gemini app (upload your survey spreadsheet first):
Help me clean my employee survey spreadsheet. Apply these rules:
1. Fill any blank values in the [name column] with "Anonymous"
2. Replace "[original label]" in the [region column] with "[standardised label]" (apply this to all variants: "[variant 1]", "[variant 2]")
3. Remove any rows where the [satisfaction column] is blank
Generate a new cleaned file and summarise how many changes were made in each category.
Fill in the variables:
[name column]— the exact column header for respondent names (e.g. "Name", "Respondent")[region column]— the column containing location/region labels (e.g. "Region", "Office")[original label]/[standardised label]— e.g. replace "Headquarters" with "HQ"; add more replacement rules as needed[variant 1],[variant 2]— other non-standard spellings to catch (e.g. "HQ Office", "Head Office")[satisfaction column]— the column that must have a value to keep the row (e.g. "Satisfaction Score", "Rating")
Customisation tips
- Add more normalisation rules: "Replace 'N/A' and '-' with blank in the department column"
- Ask for a change log: "Create a second sheet listing every row that was modified and what changed"
- Validate score ranges: "Flag any satisfaction score outside 1–5 in a new 'Flagged' column"
- Request deduplication: "Remove duplicate rows where email address matches"
Iteration suggestions
After Gemini produces the clean file:
- Summarise for leadership → use
survey-results-leadership-summaryto draft the executive email with findings from the cleaned data - Build charts → ask Gemini in Sheets to create summary charts from the cleaned spreadsheet
Related skills in this workflow
| Step | Skill | What it does |
|---|---|---|
| 1 | employee-engagement-survey |
Generate the survey question bank |
| 2 | anonymous-satisfaction-survey |
Draft the anonymous survey |
| 3 → You are here | survey-data-cleanup |
Clean and standardise the response data |
| 4 | survey-results-leadership-summary |
Summarise findings for senior leadership |