☐ Duplicate rows
One client/order recorded twice — inflates metrics. Fix: unique key (email/phone/ID), dedupe by it; Excel: 'Remove duplicates' by key column.
☐ Missing values
Empty cells break averages and filters. Fix: decide — drop the row, fill with zero/median, or tag as a separate 'unknown' category. Never silently leave as-is.
☐ Mixed date formats
01.02.2025 vs 2025-02-01 vs 'Feb 1' — Excel reads some as text. Fix: normalize to ISO YYYY-MM-DD; verify the date column is real dates.
☐ Numbers stored as text
'1 234' with a space, '5,6' with a comma — won't sum. Fix: replace spaces, comma→dot, then to number; currency in a separate column, not in text.
☐ Extra spaces & case
' Kyiv', 'kyiv', 'KYIV' — three categories instead of one. Fix: TRIM + case unification + a lookup of allowed values for key fields.
☐ Outliers
An order of 9,999,999 — input error or real case? Fix: inspect values beyond P1–P99; don't auto-delete — first understand the cause.
☐ Inconsistent categories
'FOP', 'fop', 'FOP (group 2)' — fragment the slices. Fix: validation dropdown at source; a mapping column for legacy data.
☐ Hidden characters
Non-breaking spaces, line breaks in cells — break JOINs/search. Fix: CLEAN/TRIM, regex for non-alphanumerics in key fields.