Business-friendly data models

Graham Witt

If a system is to support an enterprise’s business information requirements, a necessary part of the design process is effective review of the design by appropriate business stakeholders. For such review to be effective, the design documentation provided to those stakeholders (“the business model”) must

  • be understandable
  • be complete, i.e. depict all information in which business stakeholders are interested
  • not depict any information in which business stakeholders have no interest (“noise”) which distracts or confuses reviewers, reducing review effectiveness.

Logical data models do not meet these criteria, yet most conceptual data models are degenerate logical data models including much noise. This presentation details what should be included in and excluded from a data model to be reviewed by business stakeholders.


Graham has over 30 years of experience in delivering effective data solutions to the government, transportation, finance and utility sectors. He has specialist expertise in business requirements, architectures, information management, user interface design, data modeling, database design, data quality and business rules. He has spoken at conferences in Australia, the US, the UK, and Europe, and delivered data modeling and business rules training in Australia, Canada and the US. He has written two textbooks published by Morgan Kaufmann: “Data Modeling Essentials” (with Graeme Simsion) and “Writing Effective Business Rules”, and has written two series of articles for the Business Rule Community (www.brcommunity.com).

Comments are closed.