A missing value is an absent entry where you expected information. The absence may mean unknown, not applicable, not yet recorded or deliberately withheld. Those meanings lead to different decisions. A blank is not automatically a zero, a “no” or an error.
First check how the source system represents missing information. Then distinguish the meanings that matter for your task. Filling gaps simply to make a table look complete can replace an honest uncertainty with a false fact.
A small-business example
A supplier register has blank delivery lead times. Some suppliers are new and have no delivery history; others supply services for which delivery days do not apply. One older supplier has a value missing because a form was skipped.
Replacing all blanks with zero would imply immediate delivery. The buying team records the reason for the gap where it can be established, follows up on missing entries and excludes inapplicable records from the particular lead-time comparison with that choice clearly stated.
Try this
Choose an important column and inspect several blank entries in their source context. Ask why the value is absent. Group only the cases whose meaning you can establish. Leave the rest explicitly unknown, and decide whether the remaining uncertainty prevents the question from being answered.
Use Data dictionary to define missing-value conventions, Ask a better question to assess their relevance and Human review when the records cannot settle the meaning.