Skip to content
E
ERPResearch

What is Data Cleansing?

Data cleansing is the process of correcting, de-duplicating, and standardizing legacy data before migrating it into a new ERP system.

Definition

Data cleansing (also called data scrubbing) is the activity of identifying and fixing inaccurate, incomplete, duplicated, or inconsistently formatted records so they are fit to load into a new ERP system. It typically covers master data such as customers, suppliers, materials, items, and chart of accounts, as well as open transactional data like outstanding invoices and orders. Cleansing involves removing duplicate records, completing missing fields, standardizing formats and units, and retiring obsolete entries. Because an ERP system amplifies the value of clean data and the cost of bad data, cleansing is a critical and often underestimated workstream. Poor data quality is a frequently cited cause of ERP go-live problems and user distrust of the new system.

How Data Cleansing Works in ERP

Data cleansing usually runs in parallel with configuration, ahead of one or more migration trial loads into the test environment. Business data owners decide which legacy records to keep, transform, or archive, applying validation rules that match the new ERP's required fields. Cleansed data is then loaded for UAT so users test against believable records, and migration is repeated until load error rates are acceptable for cutover.

ERP Vendors with Strong Data Cleansing

Frequently Asked Questions

When should data cleansing start in an ERP project?

Cleansing should start early, ideally during or right after the blueprint or design phase, because it is time-consuming and depends on business data owners who have day jobs. Starting late is a common reason go-live dates slip, since dirty data cannot be loaded reliably.

Is data cleansing a one-time activity?

The bulk cleansing happens before migration, but maintaining data quality is ongoing. Many organizations establish data governance roles and validation rules in the new ERP so master data does not degrade again after go-live.

Related Terms