The segmentation filter returned 800 contacts instead of the expected 5,000. The missing 4,200 had no value in the "industry" field. They weren't excluded on purpose — the field was never collected. That's the problem with missing values: they're invisible until something breaks.
Detecting missing values — and understanding the patterns behind them — is one of the most important steps in any data quality assessment.
Types of Missing Values
Not all missing values are the same, and the cause determines the fix:
Sohovi profiles every column in your dataset for completeness and flags the exact rows where values are missing — free to try.
True nulls — The field was left empty because no value was collected. Common in optional form fields and manual data entry.
Placeholder nulls — "N/A", "Unknown", "None", "0" used to represent missing information. These look like values but aren't. A count of "not null" values will include them, hiding the true completeness problem.
Whitespace nulls — A space or invisible character in a field that appears empty but isn't technically null. These are particularly common in imported data.
System defaults — A default value (like "01/01/1900" for dates or "0" for numeric fields) automatically populated when no real value was entered.
How to Find Them
For true nulls: count the null/empty values in each column. Any standard data profiling tool or even a spreadsheet COUNTBLANK function handles this.
Sohovi runs a full data profile on any CSV or spreadsheet in under a minute — completeness rates, type distributions, outliers, and more.
For placeholder nulls: look at the most common values in each column. If "N/A" appears in the top 5 values for a field, you have hidden missing data.
For whitespace nulls: check the character count of fields that appear empty. A field with 1 character that looks blank probably contains a space.
For system defaults: check for suspiciously common values in fields that should have varied values (e.g., hundreds of records all showing "01/01/1900" as a birth date).
Understanding the Pattern
Once you've identified missing values, analyze the pattern:
Random missing — Values are missing in no particular pattern. Usually from optional fields or inconsistent entry.
Systematic missing — Values are consistently missing from a specific source, time period, or import. This points to a process problem (e.g., a form that didn't capture a field before a certain date).
Correlated missing — A field is consistently null when another field has a specific value. Understanding correlations helps determine whether the missing data is truly missing or correctly absent.
Sohovi detects all forms of missing values — including placeholder nulls and whitespace — when you profile your CSV, showing you the true completeness picture for each column.
Understanding why data is missing is as important as knowing that it is. The fix for random missing data (improve data entry) is different from the fix for systematic missing data (fix the source process).
