ERP data quality: why it decides projects

ERP data quality means: master data is unambiguous, complete and has an owner. It is the foundation of every reliable report, every migration and every AI capability in the system. That is exactly why it decides ERP projects: bad data makes every consolidation sluggish, every migration expensive and every automation unreliable. This page puts the topic in order: how poor data quality shows up in everyday work, why a system change will not repair it, in which order to cleanse and who owns data quality.

How poor data quality shows up in everyday work

Rarely as an error message, almost always as friction: the same report shows different numbers depending on the department, a customer exists three times with three payment terms, the month-end close drags because figures are reconciled by hand. Each of these frictions is a data problem dressed up as a process problem. Whoever knows the symptoms in the close usually finds the cause in the master data.

Why the new ERP will not rescue your data

A system change changes the filing, not the content. Ambiguous items stay ambiguous, just in a new interface. The expectation that the migration will sort the data „on the side“ is one of the most expensive mistakes in ERP projects. Why that is and what helps instead is covered in depth in the answer „The ERP will not rescue your data“.

Cleansing before the migration: the order

Before a migration you decide, you do not copy: which data objects come along, which get cleansed, which stay in the archive. Cleansed master data and open items always, transaction history only after honest deliberation. The approach with priorities and ownership is covered in the answer on cleansing master data before the ERP migration, and the data's path into the new system is described in Data migration to Business Central.

Who owns data quality?

The business, not IT. IT provides tools and rules, but whether a customer exists once or three times is decided by the process that creates it. In DGP projects, every central data object therefore gets a named owner on the business side before anything is migrated. Without this step, the sprawl simply keeps growing after go-live. The role question in detail is answered in „Who is responsible for ERP data quality?“.

What AI in the ERP has to do with data quality

Everything. Copilot and agents in Business Central work on your data, not on an ideal. On ambiguous master data, AI automates the errors, just faster. That is why the data foundation is the first homework before any AI initiative in the ERP, as the answer on the data foundation for AI agents shows. The good news: the same work pays twice, because clean data improves reports and processes immediately, and AI value comes on top.