CSV validation
Flag suspicious CSV values before you import
Learn why suspicious CSV values should be flagged for review instead of silently guessed or overwritten.
What this problem usually means
A CSV can be structurally valid and still contain values that break an import or corrupt the destination data. Suspicious value flagging is the layer after file repair: it looks for values that do not match the likely column type, such as invalid emails, impossible dates, non-numeric balances, mixed boolean formats, and unusual category values. The goal is to make risky values visible before import, not to pretend the tool always knows the right answer.
Why it happens
- Exports often combine data from different systems with different date, number, and boolean formats.
- Manual spreadsheet edits can introduce values like missing emails, text in number columns, or inconsistent status labels.
- Some values are ambiguous by design, such as 01/02/2026, which can mean different dates depending on the source locale.
What ImportFix checks
- Invalid-looking emails in columns that appear to contain email addresses
- Impossible or mixed date values, including dates that need user confirmation
- Non-numeric values inside mostly numeric columns
- Mixed boolean and category values that may map badly in the destination app
Common repairs
When to review manually
ImportFix should not guess when the value could reasonably mean more than one thing. A date like 04-12-2026, a status like unknown, or a phone number without a country context may be valid in one workflow and wrong in another. Flagging protects the data by showing the risk and letting the user decide.
Related guides
Keep exploring the most common import problems, or jump into the repair workspace directly.