Blank identifiers
Customer IDs, emails, or SKUs missing on rows that otherwise look finished.
Missing & inconsistent values
Missing data is not always wrong. Find out what needs attention.
A blank can mean not applicable, not yet known, or a broken export. “USA” and “United States” can be the same place written twice. The work is to investigate context before you fill, merge, or delete.
Do not treat every empty cell as a defect. Do not treat every spelling variant as a duplicate.
The actual problem
Inconsistency is often a process leaking into the file
Missing and inconsistent values are easy to see and hard to interpret. They often mark where two teams, two systems, or two time periods used different rules.
A complete-looking column of countries can still split reporting. A mostly-filled owner field can hide the rows that never received a handoff. Zeros can mean none, unknown, or a failed conversion from blank.
If you “fix” those rows without context, you can invent certainty the business never had.
What can go wrong
Patterns that deserve review
Category spelling drift
Closed Won, closed-won, and Complete used as if they were the same status.
Place and unit variants
UK / United Kingdom, kg / lb, or currency symbols mixed into numeric fields.
Date conventions mixed in one column
Month/day/year next to ISO dates, with no label for which is which.
Fields populated only sometimes
A column that is required in practice but empty for an entire source or time range.
Contradictions across columns
Country does not match phone prefix; status is Closed and the close date is blank.
Zeros that are not quantities
A zero that replaced a blank during export, or a numeric default that means “we do not know.”
Example scenario
A region field that looks populated
Operations filters a customer list by region before a campaign. The filter under-counts the North because half the North accounts are labeled differently — or not at all.
| Account | Region | Country | Employees |
|---|---|---|---|
| Harbor Pack | North | United Kingdom | 42 |
| Cedar Line | NORTH | UK | |
| Pine & Co | United Kingdom | 0 | |
| Millwright | EMEA | United Kingdom | 38 |
Filling every blank with “North” because the country is UK would be a guess. Leaving EMEA mixed with North without a mapping rule would keep the report wrong. Investigation first: which values are aliases, which are a different grain, and which blanks are unknown?
What to investigate
Separate empty, unknown, and conflicting
- Is the blank allowed by the process, or did a join/export drop the field?
- Are two labels aliases for one concept, or two real categories?
- Do other columns confirm or contradict the value you are about to standardize?
- Is a zero a quantity, a missing value, or a converted null?
- Should the record be completed, left unknown, or flagged for a human who owns the source system?
Where Auditere fits
Surface the records that need a person
Auditere helps investigate structured data: missing information, inconsistent values, suspicious combinations, and conflicts across fields — then review them before they move downstream.
It will not declare every blank wrong. That would be as careless as ignoring them. The product is built for operators who need to see candidates in context and keep control of the resolution.
Related reading: finding bad data in Excel and why quiet spreadsheet damage is expensive after import.
Investigate blanks and contradictions in context
Start a trial when you need a review workflow for missing and inconsistent structured data — not a bulk fill of empty cells.