Eliminating Master Data Errors Using Poka-yoke (Mistake Proofing) Principles in Workflows

Something is awry with the Master Data process if your organization routinely fixes Master Data errors in different business systems (ERP, CRM, Supply Chain, etc.) or, worse, lets bad Master Data adversely impact downstream processes related to Sales, Customer Service, Manufacturing, and Supply Chain Management.

It is well known in the quality world that preventing quality issues is better than fixing them. Poka-yoke is a Japanese term for mistake-proofing, and it is widely practiced in Manufacturing and Product Design.

Applying Poka-yoke principles is the ideal way to prevent bad Master Data from being created and propagated throughout the organization (in ERP, CRM, Customer Service, Supply Chain, Finance, and other systems).

Master Data Error Propagation

Errors are generally introduced into Master Data processes and systems when stakeholders unknowingly provide inaccurate information and are not in a position to evaluate the correctness of Master Data when creating, changing, or reviewing it.


Error-proofing with Poka-yoke

Poka-yoke principles ensure that stakeholders cannot provide inaccurate information at the start and help reviewers catch errors before Master Data is committed to enterprise systems for widespread use.

Methods to Error-proof Master Data

Since ZMDM primarily uses workflows for Master Data Management, we found the following as the best ways to assist stakeholders and to error-proof Master Data

  • Default Values (read-only)

Not only do default values (that are not changeable) reduce the amount of information that stakeholders have to provide, they also eliminate accidental errors

  • Drop-downs (restricted set of values)


  • Accurate Calculated Values based on rules and user selections

The example below the Material description is calculated based on values selected elsewhere in the New Material Request. These types of calculated values reduce errors and as well as improve consistency.



  • Attribute validation (required, min, max, type, size…)

In the example below the workflow enforcing the rule that Material Description cannot exceed 40 characters