Techment — Site Header

What is Data Validation

Definition

Data validation is the process of checking incoming or stored data against defined rules, types, ranges, and formats to confirm it is correct and usable, rejecting or flagging records that fail so that bad data is stopped before it reaches downstream consumers.

« Back to Glossary Index

Blocks malformed or out-of-range records at the boundary before they corrupt analytics
Encodes business rules as explicit, testable checks rather than tribal knowledge
Cuts pipeline incidents by failing fast on bad input instead of producing wrong results
Generates a clear record of what was rejected and why for troubleshooting

Real World Example

An online retailer validates every order record at ingestion, rejecting entries with negative quantities or missing SKUs into a quarantine table that the data team reviews each morning.

FAQs

What is the difference between validation and cleansing?

Validation checks whether data meets the rules and flags failures, while cleansing actively corrects or removes the problematic data.

Where should validation occur?

Ideally as early as possible, at ingestion or the producer side, so bad data is caught before it propagates.

What kinds of rules are validated?

Type, format, range, required-field presence, referential integrity, and custom business constraints are common validation rules.