Excel
Excel data cleaning checklist
Cleaning is not cosmetic. Most wrong analyses trace back to one of the six problems below.
Updated September 20, 2026
The six problems
| Problem | How to detect | Fix |
|---|---|---|
| Trailing spaces | LEN vs LEN(TRIM()) | Trim every text column |
| Numbers as text | ISNUMBER on a sample | Strip separators and convert |
| Duplicate rows | COUNTIFS on the key | De-duplicate on the key, not the whole row |
| Inconsistent categories | Unique value list | Map to a controlled list |
| Mixed date formats | Sort and inspect ends | Convert to ISO dates |
| Blanks that mean zero | Count blanks per column | Fill deliberately, and document it |
Rules for cleaning safely
- Never clean in place over the only copy of the raw data
- Prefer repeatable steps over manual edits so the next file is quick
- Record every assumption, especially blank handling
- Re-profile after cleaning to confirm what changed
Frequently asked questions
Should blanks become zero?
Only when a blank genuinely means zero. If it means unknown, filling with zero biases averages and totals.
Related reading
Analyze your data with AI
Open the GridMind workspace, bring in a CSV, Excel file or your Power BI model, and let AI build the formulas, charts and dashboards for you.